New Insights on the Uptake and Trafficking of Coenzyme Q

Abstract

Coenzyme Q (CoQ) is an essential lipid with many cellular functions, such as electron transport for cellular respiration, antioxidant protection, redox homeostasis, and ferroptosis suppression. Deficiencies in CoQ due to aging, genetic disease, or medication can be ameliorated by high-dose supplementation. As such, an understanding of the uptake and transport of CoQ may inform methods of clinical use and identify how to better treat deficiency. Here, we review what is known about the cellular uptake and intracellular distribution of CoQ from yeast, mammalian cell culture, and rodent models, as well as its absorption at the organism level. We discuss the use of these model organisms to probe the mechanisms of uptake and distribution. The literature indicates that CoQ uptake and distribution are multifaceted processes likely to have redundancies in its transport, utilizing the endomembrane system and newly identified proteins that function as lipid transporters. Impairment of the trafficking of either endogenous or exogenous CoQ exerts profound effects on metabolism and stress response. This review also highlights significant gaps in our knowledge of how CoQ is distributed within the cell and suggests future directions of research to better understand this process.

1. Introduction

Coenzyme Q (CoQ), or ubiquinone, is a vital lipid needed for generation of energy and a crucial antioxidant that preserves membrane structure and function. Although it is synthesized within the mitochondria of eukaryotes, it is present in all cellular membranes and is also a component of lipoproteins. CoQ₁₀ (the isoform synthesized by humans and most mammals) is an enormously popular supplement and a medically important therapeutic. Yet, its insolubility in water means it must be trafficked between membranes via lipid vesicles, micelles, chaperones, or transport proteins. The pathways and mechanisms that move this lipid from one membrane to another are still largely undefined. Elucidating the pathways of this trafficking will aid our understanding of the functional roles of CoQ and may lead to improved therapeutics.

In this introductory section, we review the structure and function of CoQ, discuss the genes necessary for its biosynthesis, and provide a short overview of CoQ deficiencies and therapeutic strategies. In the sections that follow, we aim to summarize the work in several model organisms used to investigate how CoQ is trafficked. We will discuss how the yeast Saccharomyces cerevisiae has been used to identify proteins required for the uptake and trafficking of CoQ₆ (the isoform of CoQ synthesized by S. cerevisiae), both to and from mitochondria. Mammalian cell culture models have also been used to identify pathways involved in the assimilation of exogenous CoQ₁₀. We will then address how studies in mice, rats, and guinea pigs have been used to identify the uptake and assimilation of exogenous dietary CoQ isoforms.

Human studies and clinical trials addressing the bioavailability and pharmacokinetics of CoQ₁₀ supplements and formulations are the topics of many extensive reviews, and are not discussed here. For readers interested in these topics we recommend the following reviews: [1,2,3,4,5,6,7,8,9]. Additionally, we will not cover the biosynthesis of CoQ in eukaryotic organisms such as Schizosaccharomyces pombe, Arabidopsis thaliana, Caenorhabditis elegans, and in microbes such as Escherichia coli as there are excellent reviews that cover these topics [10,11,12].

1.1. Structure and Function of CoQ_n

CoQ is a redox-active lipid composed of a fully substituted quinone ring and polyisoprenyl tail of variable length that is species dependent (CoQ_nwhere n is the number of isoprene units). Humans and most mammals synthesize CoQ₁₀, mice and rats produce CoQ₉ with minor amounts of CoQ₁₀, C. elegans CoQ₉, E. coli CoQ₈, and S. cerevisiae CoQ₆ [12,13,14,15]. CoQ can be found in the oxidized state, the semi-reduced state (ubisemiquinone, CoQH·), or the fully reduced state (ubiquinol, CoQH₂). CoQ isoforms with polyisoprene tails six units or longer are generally considered to reside at the midplane of the bilayer [16]; however, there are conflicting interpretations regarding the orientation of CoQ isoforms in lipid bilayers [17]. Less hydrophobic isoforms such as CoQ₂ and CoQ₄ readily diffuse across non-contiguous membranes. Meanwhile longer isoforms, including CoQ₆, CoQ₉, and CoQ₁₀, are unable to move between lipid vesicles in the absence of a transporter [18,19].

Mitochondrial CoQ plays an essential role in aerobic respiration and can be reduced by oxidoreductases involved in several metabolic pathways (Figure 1a). At the inner mitochondrial membrane, CoQ shuttles electrons along the electron transport chain by accepting electrons from NADH:ubiquinone oxidoreductase (Complex I) and succinate dehydrogenase (Complex II), and donating them to the cytochrome bc₁ complex (Complex III). Cytochrome c then transports electrons from Complex III to cytochrome oxidase (Complex IV), which directly reduces O₂ into water. The free energy from this electron transfer is used to transport protons from the matrix side to the intermembrane space, ultimately creating a proton-motive force used by ATP synthase to generate ATP [20,21,22]. Mitochondrial CoQ is also reduced by several other dehydrogenases and thus participates in diverse metabolic pathways, including sulfide:quinone oxidoreductase (SQR) for catabolism of sulfide, dihydroorotate dehydrogenase (DHODH) for pyrimidine biosynthesis, choline dehydrogenase (CHDH) for choline oxidation, glycerol-3-phosphate dehydrogenase (GPDH) for the glycerol-3-phosphate shuttle, proline dehydrogenase (PRODH) for catabolism of proline, and electron-transferring-flavoprotein dehydrogenase (ETFDH) for oxidation of fatty acids, branched-chain amino acids, sarcosine, and dimethylglycine (Figure 1a) [23].

Figure 1.

Coenzyme Q functions. (a) Mitochondrial CoQ shuttles electrons in the electron transport chain and is reduced by dehydrogenases in diverse metabolic pathways. CI, Complex I; CII, Complex II; CIII, Complex III; CIV, Complex IV; CV, Complex V; DHODH, dihydroorotate dehydrogenase; GPDH, glyceraldehyde-3-phosphate dehydrogenase; SQR, sulfide:quinone oxidoreductase; CHDH, choline dehydrogenase; PRDH, proline dehydrogenase; ETF, electron transfer flavoprotein; ETFDH, electron transfer flavoprotein dehydrogenase. (b) CoQH₂ is present in cellular membranes and lipoproteins and acts as a chain-terminating antioxidant to inhibit both the initiation and propagation steps of lipid autoxidation. CoQH₂ or ascorbate regenerate vitamin E from the tocopheroxyl radical. (c) CoQ functions in the plasma membrane redox system (PMRS) with other antioxidants, such as vitamin E and ascorbate. CoQ is reduced by NAD(P)H:quinone oxidoreductase 1 (NQO1), or NADH-cytochrome b₅ reductase (cyb5Red). (d) The CoQ reductase ferroptosis suppressor protein 1 (FSP1) regenerates CoQH₂ at the plasma membrane to terminate lipid peroxidation and suppress ferroptosis. Created with Biorender.com (accessed on 23 May 2023).

In mitochondrial and other biological membranes, CoQ functions as a powerful antioxidant and plays a role in cell maintenance. CoQ is the only endogenously synthesized lipid-soluble antioxidant and acts to inhibit lipid peroxidation from damaging cellular membranes, proteins, and DNA. CoQH₂ acts as a chain terminator to inhibit both the initiation and propagation steps of autoxidation of polyunsaturated fatty acids (PUFAs) [24,25]. Moreover, CoQH₂ is also capable of regenerating α-tocopherol (vitamin E) from the tocopheroxyl radical (Figure 1b) [26]. Recent studies indicate that the oxidized form of CoQ can also slow peroxyl radical addition reactions responsible for cellular damage incurred by electrophilic stress resulting from conjugated PUFAs, such as conjugated linoleic and conjugated linolenic acids [27].

CoQ at the plasma membrane is part of the plasma membrane redox system (PMRS), which combats oxidative stress, maintains cell bioenergetics when mitochondrial activity decreases, and is involved in cell growth and resistance to apoptosis (Figure 1c) [28,29]. The PMRS comprises NAD(P)H:quinone oxidoreductase 1 (NQO1), NADH-cytochrome b₅ reductase (cyb5Red), CoQ, and other antioxidants, such as vitamin E and ascorbate. The PMRS functions to regulate cellular redox homeostasis through the maintenance of cytosolic NAD(P)⁺/NAD(P)H and CoQ/CoQH₂ pools within the plasma membrane. CoQH₂ generated by the PMRS can function as a non-competitive inhibitor of neutral sphingomyelinase, a membrane protein that releases ceramide from membrane sphingomyelin to activate ceramide-dependent caspases and elicit apoptosis in serum-deprived cells [29,30,31]. Additionally, ferroptosis suppressor protein 1 (FSP1) was recently identified as a CoQ reductase that regenerates CoQH₂ at the plasma membrane to inhibit ferroptosis in a pathway that is distinct from the anti-ferroptotic glutathione peroxidase 4 (GPX4) system (Figure 1d) [32,33]. Intriguingly, vitamin K is also reduced by FSP1 and inhibits ferroptosis [34], while tetrahydrobiopterin may function as a co-antioxidant with CoQ to inhibit ferroptosis [35]

1.2. CoQ Biosynthesis, Mutations in COQ Genes, and Therapeutic Strategies

CoQ is synthesized at the inner mitochondrial membrane on the matrix side, but is found throughout all cellular membranes [36]. CoQ biosynthesis requires a high molecular mass complex, termed the CoQ synthome in yeast and complex Q in mammals, situated on the matrix side of the inner mitochondrial membrane (Figure 2) [37,38,39]. CoQ biosynthesis begins with formation of the aromatic ring precursor 4-hydroxybenzoic acid (4-HB). In yeast, 4-HB is derived from the shikimate pathway or from tyrosine [38]. Human cells, however, may derive 4-HB from phenylalanine or tyrosine due to the presence of phenylalanine hydroxylase [37]. Deamination of tyrosine produces 4-hydroxyphenylpyruvate (4-HPP). In humans, mouse, rat, and guinea pig, this step is thought to be catalyzed by tyrosine aminotransferase (TAT) or by mitochondrial alpha-aminoadipate aminotransferase (AATAD) [37]. In yeast there is considerable redundancy in the pathway of 4-HB synthesis; five aminotransferase enzymes can perform this deamination step [40]. In humans, hydroxyphenylpyruvate dioxygenase-like protein (HPLD) converts 4-HPP to 4-hydroxymandelate (4-HMA) [41]. 4-HMA is also a likely intermediate in the yeast pathway to 4-HB [40,42]. The additional steps producing 4-hydroxybenzaldehyde (4-Hbz) have yet to be determined in either yeast or human cells. In yeast, Hfd1 is required to oxidize 4-Hbz to 4-HB, and the human aldehyde dehydrogenase 3A1 (ALDH3A1) has also been shown to perform this step [37]. COQ2, or Coq2 in yeast, first attaches the variable length polyisoprenyl tail to 4-HB. The polyprenylated intermediate is then introduced to the CoQ synthome (or complex Q), where the COQ3-COQ9 polypeptides act in concert to synthesize CoQ (Figure 2) [10,38]. The final CoQ product is then transported to its intracellular destination in the mitochondria, plasma membrane, or other organellar membranes via poorly characterized pathways [37].

Mutations in the CoQ biosynthetic genes may result in a condition termed primary CoQ₁₀ deficiency. Primary CoQ₁₀ deficiency is a rare disease often associated with multisystem disorders [43]. However, deleterious mutations that produce a poorly functioning or non-functional polypeptide, can also result in isolated phenotypic consequences, affecting the kidney or central nervous system. Genetic or biochemical testing is required for comprehensive diagnosis [43,44]. Secondary CoQ₁₀ deficiency may arise when mutations in genes not directly associated with CoQ₁₀ biosynthesis affect cellular CoQ₁₀ content [45], in response to drugs (such as statins), or as a result of aging [39]. Currently, the only treatment for CoQ₁₀ deficiency is high-dose exogenous supplementation of CoQ₁₀. While CoQ₁₀ displays poor bioavailability, in some patients presenting with primary CoQ₁₀ deficiency supplementation produced dramatic improvements in their long-term clinical phenotypes [6,46]. To improve such clinical outcomes, it is important to understand both the gastrointestinal (GI) uptake pathway of orally supplemented CoQ₁₀, as well as the cellular uptake and intracellular distribution.

Figure 2.

Coenzyme Q biosynthetic pathway. The dashed arrows indicate decarboxylation and hydroxylation steps that are catalyzed by unknown enzyme(s), thus resulting in an uncertainty in the order of reactions. Intermediates in the pathway include 4-HPP, 4-hydroxyphenylpyruvate; 4-HMA, 4-hydroxymandelate; 4-Hbz, 4-hydroxybenzaldehyde; 4-HB, 4-hydroxybenzoic acid; DMAPP, dimethylallyl pyrophosphate; IPP, isopentenyl pyrophosphate; HPB, 4-hydroxy-3-polyprenyl-benzoic acid; DHPB, 4,5-dihydroxy-3-polyprenylbenzoic acid; HMPB, 4-hydroxy-5-methoxy-3-polyprenylbenzoic acid; DHP, 4,5-dihydroxy-3-polyprenylphenol; DDMQH₂, 2-methoxy-6-polyprenyl-1,4-benzohydroquinone; DMQH₂, 2-methoxy-5-methyl-6-polyprenyl-1,4-benzohydroquinone; DMeQH₂, 3-methyl-6-methoxy-2-polyprenyl-1,4,5-benzenetriol; n is the number of isoprene units in the polyisoprenyl tail. Yeast can also utilize p-aminobenzoic acid as a ring precursor, and its prenylation by Coq2 produces the early intermediate 4-amino-3-polyprenyl-benzoic acid (HAB, [38]). Adapted with permission from Wang et al. [44]. Created with Biorender.com (accessed on 28 June 2023).

The rescue of CoQ₁₀-deficient animals supplemented with synthetic hydrophilic CoQ precursors or “bypass therapies” indicates that bioavailability is a key limitation of CoQ₁₀ supplementation [6]. Bypass therapy for patients with primary CoQ deficiency is currently being studied as a potential alternative to supplementation with CoQ₁₀ [47,48]. In bypass therapies, analogs of 4-HB are used as alternate ring precursors that function to “bypass” a blocked step in the CoQ biosynthetic pathway. For example, 2,4-dihydroxy-benzoic acid (2,4-diHB) partially restores CoQ biosynthesis in mice harboring mutations in Coq7 (also termed Mclk1) [49]. Mice with certain mutations of Coq9 also respond to 2,4-diHB treatment [50]. However, 2,4-diHB also inhibits the endogenous CoQ biosynthetic pathway, possibly due to competition with natural substrates of the COQ enzymes upstream of the Coq7 hydroxylation step [49]. Bypass of human COQ6 mutations using vanillic acid, a methoxylated analog of 4-HB, also has efficacy in a cultured cell model [51]. Due to the lack of a hydrophobic polyisoprenyl tail, the analogs of 4-HB are significantly more bioavailable than CoQ₁₀, and the restoration of endogenous biosynthesis may achieve a more effective rescue [52].

2. The S. cerevisiae Model

S. cerevisiae (hereafter referred to as yeast) is a single-celled fungus that has been widely studied as a simple, eukaryotic model organism. Yeast cells are capable of both fermentation and aerobic respiration, and they divide rapidly through a budding process in which smaller daughter cells bud off from a larger mother cell. These cells are approximately five microns in diameter, between bacterial and human cells in size. Yeast cells can readily switch between haploid a and haploid α cells and mate to form diploid cells, which can further sporulate to produce four haploids [53]. Many eukaryotic cellular processes have been first identified and extensively characterized in yeast, including cell cycle regulation, vesicle trafficking, protein secretion, and autophagy. Many of these processes are highly conserved between yeast and humans. In fact, 47% of the 414 essential genes in S. cerevisiae with a 1:1 human ortholog can be functionally replaced by the corresponding human ortholog [54]. These well-characterized pathways, ease of genetic manipulation, and strong conservation between yeast and humans make yeast a powerful model organism.

Much work has been performed to elucidate the biosynthetic pathway and understand the cellular function of CoQ₆ in yeast. The COQ1–COQ11 genes encode mitochondrially targeted polypeptides that are required for efficient CoQ₆ biosynthesis [37,38,39]. Strains with a deletion or deleterious mutation in one of the COQ1–COQ9 genes, termed coq mutants, are unable to synthesize CoQ₆ (Figure 3a). These coq mutants are thus incapable of cellular respiration and fail to grow in a non-fermentable medium. In most cases, expression of the human COQ ortholog in the respective yeast coq mutant rescues efficient CoQ₆ biosynthesis and restores growth in a non-fermentable medium, highlighting the strong conservation of CoQ biosynthesis from yeast to humans. In addition to the presence of all the Coq polypeptides, efficient CoQ₆ synthesis requires Coq3–Coq9 and Coq11 to form a mega-complex on the matrix side of the inner mitochondrial membrane, termed the CoQ synthome [38,39]. The formation of this complex is analogous to the Complex Q assembly that occurs in human cells [39]. The strong conservation of CoQ biosynthesis and many functions from yeast to humans suggest a conserved pathway of CoQ transport.

Figure 3.

Yeast coq mutants are rescued with exogenous CoQ₆ but not non-isoprenoid CoQ analogues. (a) CoQ₆ was not detected in any of the coqΔ mutants. CoQ₆ (pmol/mg protein) was determined based on a CoQ₆ standard curve as described in (Tsui et al., 2019). (b) Each of the coq mutants, from coq3Δ to coq9Δ, accumulated early CoQ₆ intermediates (HHB and HAB). The amount of HHB and HAB was calculated as the area of the MS peak/mg protein. (c) Yeast mutants harboring deletions in either the COQ1 or COQ2 gene showed more robust rescue in response to exogenous CoQ₆ treatment than do the other single coq null mutants (coq3Δ–coq9Δ). (d) Additional deletions of either COQ1 or COQ2 restored the deficient CoQ₆-rescue of a coq3Δ mutant, but do not affect the phenotype observed in coq1Δ or coqΔ2 strains. Columns represent the degree of rescue (in %) ± SD of a strain compared to WT, which is defined as 100% and represented as a dashed line. Statistically significant differences between a specific coqΔ mutant and one of its counterparts (another coqΔ mutant) are denoted with numbers in parentheses on top of the columns. (1) represents differences comparing to coq1Δ, (2) represents differences comparing to coq2Δ, (3) represents differences comparing to coq3Δ, etc. Three or more independent rescue experiments were performed for every strain. Asterisks on top of the columns represent significant differences when compared to WT (* p < 0.05, ** p < 0.01, *** p < 0.001). (e) Structures of CoQ analogues. Panels (a–d) reproduced with permission from Fernandez-del-Rio et al. Free Radicals in Biology and Medicine, published by Elsevier, 2020 [18]. Created with Biorender.com (Accessed on 27 June 2023).

2.1. Growth Rescue of coq Mutants in a Non-Fermentable Medium Supplemented with Exogenous CoQ₆

As previously stated, coq mutants are unable to synthesize CoQ₆ and are thus respiratory deficient and fail to grow in a medium containing a non-fermentable carbon source (Figure 3a). Deletion of COQ1 or COQ2 that are involved in early steps of CoQ₆ biosynthesis prevents the accumulation of early CoQ₆ intermediates, 3-hexaprenyl-4-hydroxybenzoic acid (HHB) and 3-hexaprenyl-4-aminobenzoic acid (HAB), while deletion of any one of the COQ3-COQ9 genes that are required for later biosynthetic steps accumulate these early CoQ intermediates (Figure 3b). Significantly, direct supplementation of exogenous CoQ₆ restores respiration and growth of coq mutants [55,56], though the degree of rescue depends on the nature of the coq mutant. When supplemented with exogenous CoQ₆ in non-fermentable media, coq1Δ and coq2Δ single mutants exhibit wild-type or near wild-type levels of growth, while each of the coq3Δ–coq9Δ single mutants exhibit growth that is significantly lower than wild-type cells (Figure 3c). Deletion of either COQ1 or COQ2 in the coq3Δ single mutant restores rescue to that of wild-type cells or to the coq1Δ and coq2Δ single mutants (Figure 3d), indicating that the accumulation of early CoQ₆ intermediates, HHB and HAB, impairs growth rescue of coq3Δ–coq9Δ mutants [18]. It is possible that HHB/HAB act as detergents or may interfere with membrane trafficking steps needed for transit of exogenously added CoQ₆. CoQ₆ uptake is favored by the presence of peptone in the culture media [57]. The digested proteins in peptone likely bind CoQ₆ and enable its entry into the aqueous compartment of endocytic vesicles. Following supplementation with exogenous CoQ₆, the coq mutants also show a significant delay in growth compared to wild-type cells as they make the diauxic shift from fermentation to respiration [57]. Supplementation with either hydrophilic CoQ isoforms that can spontaneously diffuse across non-continuous membranes (CoQ₂, CoQ₄), or hydrophobic CoQ isoforms that cannot diffuse (CoQ₆) are both capable of rescuing growth. Moreover, supplementation with CoQ₂ can bypass defects in endocytosis and membrane trafficking [19,58]. The ability of yeast to uptake exogenous CoQ, transport it to the mitochondria, and in turn restore respiration in coq mutants makes yeast a powerful model organism to investigate CoQ uptake and transport.

Studies with supplementation of non-isoprenoid CoQ analogs indicate that the isoprenoid tail is essential for efficient uptake and growth rescue of coq mutants in a non-fermentable medium. Idebenone, decylQ, and MitoQ are three such analogs containing a ten-carbon alkyl tail (Figure 3e). The tail end of idebenone and MitoQ are further modified with a hydroxyl group or triphenylphosphonium moiety, respectively, with the latter exhibiting mitochondrial targeting. None of these three analogs were able to restore growth of coq mutants when supplemented in a non-fermentable medium [58]. Although decylQ and idebenone restore respiration in isolated mitochondria, it is speculated that mitochondrial accumulation of these analogs in whole cells is insufficient to support growth. Conversely, MitoQ accumulates in mitochondria of whole cells, but is poorly oxidized by Complex III and thus unable to restore respiration [58]. In fact, supplementation with high concentrations of decylQ inhibits growth rescue of coq mutants with CoQ₂, perhaps through competition with an unidentified CoQ-binding protein [19]. Supplementation assays with analogs of CoQ₂ where the prenyl tail has been modified show that each of the analogs tested are capable of restoring respiration in isolated mitochondria, but the presence of the double bond between C2 and C3 in the first isoprenoid unit of exogenously supplied CoQ is essential for efficient uptake in whole cells and growth in a non-fermentable medium [19]. Taken together, the decylQ impairment of growth rescue by CoQ₂ and the structural requirement in the isoprenoid tail of exogenous CoQ suggest the presence of an essential protein that can distinguish subtle differences in CoQ structure for efficient delivery or retention of exogenous CoQ to mitochondrial respiratory complexes.

It is possible that the yeast Coq10 polypeptide may fulfill this role of a CoQ chaperone. Yeast Coq10 is peripherally associated with the matrix side of the inner mitochondrial membrane, but is not part of the endocytic pathway, the respiratory complexes, or the CoQ synthome [59]. The yeast Coq10 polypeptide contains a steroidogenic acute regulatory-related lipid transfer (START) domain, which binds CoQ_n isoforms of variable tail lengths and analogs of late-stage CoQ-intermediates such as DMQ₃ [60]. Yeast Coq10 is needed for respiratory function, and may be involved in transport of newly synthesized CoQ₆ to Complex III [61]. COQ10A and COQ10B are human orthologs of yeast Coq10, and partially rescue both the respiratory defect and sensitivity to lipid peroxidation of the coq10Δ yeast mutant [62]. Human STARD7 is structurally related to the yeast Coq10 and human COQ10 polypeptides, binds CoQ₄ in vitro, and plays a role in vivo in transporting mitochondrial CoQ₁₀ to the plasma membrane to suppress ferroptosis [63]. Human saposin B binds CoQ₁₀, and is postulated to function as a transporter [64,65]. Mammalian CoQ transport proteins are discussed in Section 3.1.5. It is likely that many other CoQ transporter or chaperone proteins remain to be discovered.

2.2. Identification of Essential Genes in CoQ₆ Uptake and Trafficking to Mitochondria Respiratory Complexes with the Use of Yeast ORFΔcoqΔ Double Mutants

2.2.1. A Screen for Genes That Impact Growth Rescue by Exogenous CoQ

Previous studies performed by James et al. have sought to identify genes that influence the ability of exogenously supplemented CoQ₂ and CoQ₄ to support the respiratory growth of CoQ-less yeast [19]. Sixteen such genes were found by screening a library of ~4800 ORFΔcoqΔ double mutants. The deletion of these ORFs resulted in impaired growth rescue on a medium containing a non-fermentable carbon source when supplemented with exogenous CoQ₂ or CoQ₄. However, it is important to note that CoQ₂ and CoQ₄ have shorter isoprenoid tails than CoQ₆, passively diffuse across noncontinuous membranes, and move independently of lipid trafficking mechanisms [18,19].

To further examine the role of these ORFs as well as other candidates in the uptake and trafficking of CoQ₆, Fernández-Del-Río et al. tested whether exogenous CoQ₆ restored growth in a screen of 40 ORFΔcoq2Δ double mutants in a non-fermentable medium [18]. This group of 40 genes included the 16 ORFΔcoq2Δ double mutants identified by James et al. [19]. The rescue assay compared the ORFΔcoq2Δ double mutant to the coq2Δ single mutant, and to the ORFΔ single mutant in a non-fermentable medium when supplemented with exogenous CoQ₆ [18]. The degree of growth rescue in a medium supplemented with exogenous CoQ₆ was shown to be dependent on the distinct COQ gene that was disrupted (see Section 2.1). Six ORFΔcoq2Δ double mutants demonstrated reduced growth in a non-fermentable medium with exogenous CoQ₆ as compared to either the coq2Δ or the ORFΔ single mutants, suggesting a role for these ORFs in CoQ₆ uptake and transport to the mitochondria. Fernández-Del-Río et al. reasoned that supplementation with the hydrophilic isoform CoQ₂ should be able to circumvent the trafficking defect observed with CoQ₆ in these mutants [18]. Indeed, the growth rescue afforded by CoQ₂ supplementation in these ORFΔcoq2Δ double mutants far exceeded that of CoQ₆. Such growth rescue assays of ORFΔcoq2Δ double mutants provide a powerful tool to identify genes that are required for the uptake of exogenous CoQ₆ and its trafficking to mitochondria. However, more work is needed to unravel the role these gene products play in uptake and transport of exogenous CoQ₆.

Interestingly, two genes, RTS1 and CDC10, were identified as essential in both screens (Figure 4) [18,19]. Rts1 is a regulatory subunit of protein phosphatase 2A (PP2A), which plays a role in regulation of cell size control, the mitotic spindle orientation checkpoint, and septin ring organization and disassembly during cytokinesis [66,67]. It is possible that Rts1 in complex with PP2A plays a regulatory role in CoQ₆ uptake and transport. CDC10 is one of four core septins in yeast, a group of GTPases that assembles into filaments and higher order structures [68]. Septin filaments assemble into a septin ring at the bud neck in yeast and act as a scaffold to recruit other proteins to drive cytokinesis [69]. Septins have been shown to interact with a subset of endocytic proteins [70], and SEPT9, the human homolog of Cdc10, has been implicated in endosomal trafficking [71]. More work is needed to determine if the role of Cdc10 in CoQ₆ uptake and transport is involved in endocytosis or another septin function.

Figure 4.

Pathways and proteins involved in trafficking of CoQ₆ to and from mitochondria. See text for the description of gene products that are involved in transport of CoQ₆ in yeast. The endoplasmic reticulum-mitochondria encounter structure (ERMES) is formed by Mdm10 and Mdm34 at the outer mitochondrial membrane, Mdm12 in the cytosol, and Mmm1 on the ER membrane. Septin filaments are hetero-oligomeric complexes composed of septin subunits, including Cdc10, Cdc3, Cdc12, Cdc11, and sometimes Shs1. PP2A, protein phosphatase 2A. Dashed lines denoted with question marks indicate postulated pathways of CoQ trafficking. Created with Biorender.com (Accessed on 25 May 2023).

The four other essential genes for uptake and trafficking of exogenous CoQ₆ to respiratory complexes identified by Fernández-Del-Río et al. are RVS161, RVS167, NAT3, and VPS1 (Figure 4) [18]. Rvs161 and Rvs167 are amphiphysin-like membrane proteins that form a heterologous dimer and promote membrane curvature [72]. Rvs161 and Rvs167 interact with several other proteins, including Vps1, at the plasma membrane for the vesicle scission step of endocytosis [73]. Vps1 is a GTPase involved in membrane fusion and fission. Vps1 also operates in other cellular processes, including vacuolar protein sorting, actin cytoskeleton organization, late Golgi retention of proteins, and peroxisome biogenesis [74]. Nat3 is a catalytic subunit of the NatB complex that mediates acetylation of the N-terminus of approximately 20% of proteins in yeast and humans [75,76]. The role that each of these genes plays in CoQ₆ uptake and trafficking remains unclear.

Typically, mitochondria are excluded from schemes depicting endosomal trafficking [77]. This is curious because trafficking between the plasma membrane and mitochondria is reported to occur via clathrin-mediated endocytosis [78]. However, trafficking of proteins and lipids to mitochondria is likely mediated by extensive contact sites between mitochondria and other intracellular membranes (Figure 4) [79,80]. Fernández-Del-Rio et al. screened ORFΔcoq2Δ double mutants predicted to impair selected membrane contact sites such as the endoplasmic reticulum-mitochondria encounter structure (ERMES), the nucleus-vacuole junction (NVJ), and the vacuole–mitochondria patch (vCLAMP) [18].

However, none of these ORFΔcoq2Δ double mutants displayed impaired growth rescue when treated with exogenous CoQ₆. These results could suggest that CoQ₆ transport may occur independently of these membrane contact sites, or that transport through contact sites is redundant. For example, mitochondria dependence on phospholipid exchange has been shown to depend on either ERMES or vCLAMP [81]. Loss of ERMES stimulates vCLAMP and vice versa; however, loss of both is lethal [81]. Yeast genes that are redundant or essential for growth on a non-fermentable carbon source will not be detected by such exogenous CoQ₆ growth-rescue screens. It is also possible that other lipid transport mechanisms operate independently of both endocytic and contact site-mediated trafficking.

2.2.2. Rescue of coq Mutants with Exogenous CoQ Is Variable and Depends on the Genetic Background of the Yeast Strain

Growth rescue of coq mutants in a non-fermentable medium with exogenous CoQ₆ is dependent on the yeast genetic background. For example, the commonly used BY4741/BY4742 laboratory yeast strain takes up and transports exogenous CoQ₆ to the mitochondrial respiratory complexes, and was used in the screen to detect trafficking defects (Section 2.2.1) [18]. This is also the case for the W303-1B and CEN.PK2-1C yeast genetic backgrounds [82]. Yet, certain yeast strains harboring a coq7Δ null allele, such as the EG103coq7Δ or the FY250coq7Δ mutant yeast strains, fail to be rescued when the growth medium is supplemented with exogenous CoQ₆. The content of CoQ₆ in the gradient-purified mitochondrial fraction of these strains was only 35% and 8% of the content present in the mitochondria isolated from the EG103 and FY250 wild-type parental strains, respectively. However, the CoQ₆ content in the plasma membrane fraction isolated from EG103coq7Δ and the FY250coq7Δ strains exceeded or equaled the content of CoQ₆ normally present in the plasma membrane fraction of the parental strains [82]. In these mutant strains, the plasma membrane redox activity was augmented, suggesting that the exogenous CoQ₆ present in the plasma membrane was functional [82]. Thus, it is likely that such strains unable to be rescued for growth by exogenous CoQ₆ likely have a defect in trafficking the CoQ₆ from the plasma membrane to the mitochondria [82]. Both the parental EG103 and the EG103coq7Δ mutant strains were shown to secrete carboxy-peptidase Y and accumulate small vacuoles, markers that characterize impaired membrane trafficking [57]. It would be important to test whether these mutants are rescued by supplementation with exogenous CoQ₂, since this analog does not depend on lipid trafficking.

To test the idea that defects in endocytosis or endomembrane trafficking alter CoQ₆ uptake and transport to the mitochondria, Padilla-López et al. interrogated ORFΔcoq3Δ double mutants whose ORF deletion, such as ERG2, PEP12, TLG2, or VPS45, disrupts several endocytic transport steps [57]. Supplementation with exogenous CoQ₆ failed to rescue growth of these ORFΔcoq3Δ double mutants in a non-fermentable medium, suggesting that the endocytic and endomembrane trafficking pathways are involved in CoQ₆ uptake and transport. Subcellular fractionation confirmed that each of these ORFΔcoq3Δ double mutants have decreased uptake of exogenous CoQ₆ in mitochondria, Golgi, vacuole, and plasma membrane compared to WT and coq3Δ controls, further illustrating the impact of disrupting these trafficking pathways in CoQ₆ uptake. Moreover, the low CoQ₆ content in the plasma membrane of these mutants suggests that most of the exogenously supplied CoQ₆ cannot be directly incorporated into the plasma membrane.

It must be noted that both coq7Δ and coq3Δ mutants accumulate HHB and HAB, early CoQ₆ intermediates that act to impair growth rescue in a non-fermentable medium with exogenous CoQ₆. Indeed, Fernández-Del-Río et al. showed that these ORFΔcoq3Δ double mutants had significantly less robust growth in a non-fermentable medium when compared to the ORFΔcoq2Δ double mutant counterpart (Figure 3c,d) [18]. More work must be performed in yeast to further determine the role of endocytosis and endomembrane trafficking in CoQ₆ uptake and transport in yeast models. Previous work in the human promyelocytic HL-60 cell line has shown that both endogenously synthesized CoQ₁₀ and exogenously supplied CoQ₉ are distributed across membranes through the brefeldin A-sensitive endo-exocytic pathway (see Section 3.1.1) [83]; however, such experiments have not been performed in yeast.

2.3. CoQ₆ Transport to and from the Mitochondria

It is likely that uptake and trafficking of exogenous and endogenous CoQ₆ share common pathways. Recent studies suggest that the ERMES may play a role in both CoQ₆ biosynthesis and distribution. ERMES is a tethering complex that forms at the membrane contact site between the ER and the outer mitochondrial membrane and is composed of two mitochondrial subunits embedded in the outer membrane (Mdm10 and Mdm34), an ER-localized subunit (Mmm1), and a cytosolic subunit (Mdm12) (Figure 4) [84]. The four subunits are functionally linked to phospholipid biosynthesis and calcium signaling [84]. Recent structural analyses show that ERMES forms bridge-like structures between the ER and mitochondrial outer membrane, consistent with providing a conduit for phospholipid transport [85]. The mitochondrial contact site and cristae organization system (MICOS), a mega complex that regulates cristae junction organization, has been shown to assemble near ERMES sites, and ERMES function is required for MICOS assembly [86]. The Mmm1-Mdm12 complex has been shown to mediate phospholipid transfer in vitro and mutations in MMM1 and MDM12 lead to impaired lipid transfer through ERMES in vivo [87].

Intriguingly, Coq polypeptides that are components of the CoQ synthome at the inner mitochondrial membrane are visualized as puncta or CoQ domains that lie adjacent to ERMES puncta at the outer mitochondrial membrane [88,89]. Disruption of ERMES decreases the number of CoQ domains [88,89]. Moreover, ERMES mutants show increased whole-cell accumulation of CoQ₆ and CoQ₆-intermediates but decreased content of CoQ₆ in isolated mitochondria, suggesting that CoQ₆ and CoQ₆-intermediates accumulate in non-mitochondrial membranes [89]. Current observations suggest that destabilization of the CoQ synthome results in reduced sequestration of CoQ₆ and CoQ₆ intermediates in the mitochondria; alternatively, disruption of ERMES might lead to enhanced stability of non-mitochondrial CoQ₆ [89]. Despite this role for ERMES in CoQ domain formation and CoQ₆ production and distribution, direct transport of CoQ₆ through ERMES has not yet been observed. Moreover, none of the four ERMES genes were identified as essential for CoQ uptake and transport in ORFΔcoq2Δ double mutants [18]. However, disruption of ERMES results in an expansion of vCLAMP contact sites, and deletion of both contacts has been shown to be lethal [81]. Thus, disruption of ERMES might increase compensatory contact sites to counteract potential defects in CoQ₆ trafficking.

The idea that other membrane contact sites can compensate for lipid transport defects in ERMES mutants has been reinforced by recent structure/function studies of the VPS13 family of proteins. Yeast Vps13 is a 360 kDa protein composed of an extensive beta-sheet or “taco shell” lined by hydrophobic residues that can form a conduit for phospholipid transfer between organelles [90]. Yeast Vps13, Vps39, and Mcp1 form a mitochondria–vacuole contact site (Figure 4), and are required for the survival of yeast ERMES mutants [91]. Mcp1 is a mitochondrial outer membrane phospholipid scramblase, and partners with the Vps13 lipid bridge protein to re-equilibrate lipids as they are either inserted into or removed from the cytosolic leaflet of acceptor or donor membranes [92]. Vps39 is the vacuole fusion factor responsible for vCLAMP contact sites [81]. It is not yet known whether these bridge-like phospholipid transporters strategically positioned at contact sites also move neutral lipids, such as CoQ.

Recent work has identified yeast mitochondrial Cqd1 and Cqd2 proteins that influence cellular distribution of endogenous CoQ₆ [93]. Intriguingly, both are inner membrane proteins that face the intermembrane space, and potentially provide a connection for transfer of CoQ₆ from the inner membrane to the outer membrane and the ER (Figure 4). Disruption of CQD1 or CQD2 either diminishes or enhances the mitochondrial CoQ₆ content, respectively, without altering overall cellular CoQ₆ content [93]. Cqd1 and Cqd2 are members of the UbiB family of atypical kinases/ATPases. Coq8, an essential protein for CoQ₆ biosynthesis and part of the CoQ synthome, is also a member of the UbiB family whose ATPase activity may be coupled to the extraction of hydrophobic CoQ₆ intermediates from the inner membrane for subsequent processing by membrane-associated Coq polypeptides in the matrix [94]. Point mutations of both Cqd1 and Cqd2 at conserved protein-kinase-like (PKL) or UbiB motif residues disrupted their individual impact on CoQ₆ distribution, indicating that Cqd1 and Cqd2 activities rely on atypical kinase/ATPase activity, though further biochemical work is needed to prove enzymatic activity. Furthermore, haploinsufficiency of the CQD1 human ortholog ADCK2 is reported to cause aberrant mitochondrial lipid oxidation and myopathy associated with CoQ₁₀ deficiency [95].

Cqd1 participates in a contact site between outer and inner mitochondrial membranes (Figure 4), and overexpression of Cqd1 and Cqd2 elicits contact sites between the ER and mitochondria [96]. Taken in conjunction with the effect of Cqd1 on CoQ₆ retention within mitochondria, and the effect of Cqd2 on promoting CoQ₆ exit from mitochondria, these proteins likely play a role in CoQ₆ trafficking to and from the mitochondria in yeast. It would be interesting to determine whether the Cqd1 complex is co-localized with the CoQ synthome and ERMES, as it provides the “missing link” between these two. Interestingly, deletion of both Cqd1 and Cqd2 in yeast restores the normal distribution of yeast CoQ₆ [93], indicating that other trafficking mechanisms must operate. Hence, it is likely that other CoQ₆ transporters contribute to trafficking.

Distribution of CoQ₆ between mitochondrial and non-mitochondrial membranes plays important functional roles in mitochondrial respiration, response to oxidative stress, and redox control. The cqd1Δ single mutant has decreased mitochondrial CoQ₆ content, grows slowly in a medium containing glycerol as the sole non-fermentable carbon source, has increased non-mitochondrial CoQ₆ content, and shows enhanced resistance to treatment with PUFAs [93]. Conversely, the cqd2Δ single mutant has increased mitochondrial CoQ₆ content, normal growth on a medium containing glycerol, and is more sensitive to treatment with PUFAs [93]. The trafficking of CoQ to non-mitochondrial membranes (including the plasma membrane) mediates resistance to ferroptosis, a type of cell death due to lipid peroxidation [32,33]. CoQ₆ in the plasma membrane functions as a potent antioxidant and co-antioxidant (Figure 1b), and is also in the PMRS (Figure 1c). In respiratory deficient cells (yeast and HL-60 cultured cells), there is an increased content of CoQ at the plasma membrane [97,98]. This recruitment of CoQ to the PMRS enables cytosolic NADH to be oxidized to NAD⁺, together with export of the electrons to extracellular acceptors, such as ascorbate, and thereby maintains the ratio of cytosolic NAD⁺/NADH [99,100]. Hence, the pathways that control the trafficking of both endogenous and exogenous CoQ exert profound effects on metabolism and stress response.

3. Mammalian Cell Culture Models

Mammalian cell culture has been integral in the study of the uptake and distribution of CoQ for many decades. Among the reasons for using cell culture(s) to study CoQ uptake and trafficking are the high degree of specificity with which one can control their environment, the ease and accuracy of administration of nutrients, hormones, and in this case CoQ, the decreased physiological variability as compared to live animals. However, whether cells in culture accurately reflect their true physiological context is often difficult to determine.

In addition to uptake and trafficking of exogenous CoQ, cell culture has been used to study the CoQ synthome and endogenous CoQ biosynthesis [101]. The functional roles of CoQH₂ as an antioxidant have also been characterized in culture [102,103], with a specific emphasis on plasma membrane-localized CoQ as a protectant against ferroptosis [32,33,104,105]. Cells in culture have likewise proven to be a powerful model for the study of CoQ uptake and distribution. In the following section, we discuss CoQ uptake and intracellular trafficking in epithelial, endothelial, blood, dermal, and HeLa cells. Subsequently, the supplementation and efficacy of CoQ₁₀H₂ and CoQ₁₀ will be compared, before concluding with a discussion about cell receptors that mediate uptake of exogenous CoQ₁₀.

3.1. Studies of CoQ Uptake in Mammalian Cell Culture

3.1.1. Epithelial Cell Lines

Orally supplemented CoQ₁₀ is first taken up by the epithelial cells of the digestive tract [2,9,106]. However, the mechanism by which enterocytes absorb CoQ₁₀ is not well understood. Several reviews and studies state that the uptake of CoQ₁₀ occurs by a simple process of passive diffusion [2,9]. A combination of passive diffusion and receptor-mediated uptake is responsible for absorption of fatty acids [107], vitamin E [108], and cholesterol [109] in enterocytes. Niemann–Pick C1-like 1 (NPC1L1) plays an important role in uptake of dietary cholesterol; its disruption in mice results in a 50% decreased cholesterol absorption [109]. Recent studies suggest that membrane protein transporters are also important for the uptake of CoQ₁₀. The human intestinal epithelial cell line CaCo-2 provides a model for much of the recent research on CoQ absorption in the digestive tract. CaCo-2 is a colorectal adenocarcinoma cell line that is used to study transport across the intestinal epithelium, including the absorption of various hormones and small molecules such as insulin [110], anthocyanin [111], and calcitonin [112]. Among the advantages of this cell line is its formation of tight junctions, microvilli that form a brush border on the luminal or apical side, and transport proteins that allow the uptake and digestion or transport of nutrients to the basolateral side, all of which can simulate the in vivo digestion of molecules via oral supplementation (Figure 5) [113].

Figure 5.

CaCo-2 cell monolayers are used to model the uptake and transport of CoQ from the apical or lumenal side to the basolateral compartment. Created with BioRender.com (Accessed on 25 May 2023).

CaCo-2 cells can assimilate solubilized CoQ₁₀ from digested oral supplements, such as CoQ₁₀ in liposomes and micelles. Supplementation of CaCo-2 cells with CoQ₁₀-infused liposomes altered the mRNA levels of 694 genes involved in signaling, metabolism, and general transport [106]. Many of these proteins are directly regulated by CoQ, such as Caspase-3, which is inhibited by CoQ₁₀ at the plasma membrane [114]. CoQ₁₀ uptake by CaCo-2 cells is enhanced when solubilized forms of CoQ₁₀ were administered via simulated digestive micellization [115]. Eight different commercial CoQ₁₀ products were mixed with fat-free yogurt and olive oil, and subjected to gastric conditions by treatment with hydrochloric acid and porcine pepsin. Following this treatment, lipase, bile extract, and porcine pancreatin were added to simulate digestion. This digestion micellized the CoQ₁₀, which was then extracted from the aqueous layer of the resultant mixture. Of the eight CoQ₁₀ products, four exhibited greater than 60% micellization in the aqueous layer: γ-Cyclodextrin complexed-CoQ₁₀ tablet, liposomal CoQ₁₀ solution, solubilized CoQ₁₀ capsule, and a γ-Cyclodextrin/solubilized CoQ₁₀ powder termed Hydro-Q-Sorb^® [115]. γ-Cyclodextrin is a water-soluble octasaccharide of particular interest to the optimization of CoQ uptake [116]. The sugar features eight glucose subunits interlinked to form a cylindrical shape, characterized by its hydrophilic exterior and hydrophobic interior, which can encapsulate CoQ₁₀. γ-Cyclodextrin is one of many formulations of CoQ₁₀ that have been assessed for their ability to enhance its bioavailability [117]. The liposomal CoQ₁₀ and solubilized CoQ₁₀ displayed the greatest rate of micellization, but it was Hydro-Q-Sorb^® that was absorbed most by CaCo-2 cells. Hydro-Q-Sorb^® exhibited a dramatic increase in CoQ₁₀ content after a four-hour incubation when compared to incubation with the non-solubilized CoQ₁₀ powder reference, despite Hydro-Q-Sorb^® containing only 20% of the concentration of CoQ₁₀ as the reference powder [115]. The correlation between CoQ₁₀ product micellization and CaCo-2 uptake was weak, suggesting that other factors regulate CoQ₁₀ uptake besides micellar transport [115]. Regardless, there does exist a positive relation between the micellization of CoQ₁₀ and its internalization in the CaCo-2 cell model, indicating the importance of lipid dispersion in CoQ uptake.

The CaCo-2 human intestinal epithelial cell model has been used to compare the relative effects of CoQ₁₀ and CoQ₁₀H₂ uptake, as several studies postulate that CoQ₁₀H₂ is more bioavailable than CoQ₁₀ [118]. Supplementation with CoQ₁₀H₂ poses a challenge because the hydroquinone undergoes rapid autoxidation under cell culturing conditions [119]. As such, studies designed to compare treatment with CoQ₁₀H₂ versus CoQ₁₀ should monitor autoxidation of the CoQ₁₀H₂ during incubation conditions. Recently, Naguib et al. [120] found that CoQ₁₀H₂ complexed with γ-cyclodextrin remained stable in an aqueous salt solution for up to seven days, with the majority of the original CoQ₁₀H₂ not undergoing autoxidation to CoQ₁₀, demonstrating that CoQ₁₀H₂ can be supplemented with high purity in this form. Failla et al. [118] investigated the relative uptake and transport of CoQ₁₀ and CoQ₁₀H₂ across the CaCo-2 monolayer using a previously described simulated digestion method [115]. CoQ₁₀ or CoQ₁₀H₂ was subjected to “digestion” and added to the apical medium. Following a four-hour incubation, the total cell-associated CoQ₁₀ content and the total amount of CoQ₁₀ transported into the basolateral medium were determined. Both apical uptake into CaCo-2 cells, and transepithelial transport was significantly higher for the micellized CoQ₁₀H₂ as compared to the micellized CoQ₁₀ [118]. The authors showed that incorporation of CoQ₁₀H₂ into micelles during the digestion process was twice as efficient when compared with CoQ₁₀ [118]. Non-complexed CoQ₁₀ has been shown to have very low permeability across the CaCo-2 monolayer [118,121], indicating that in the absence of CoQ₁₀ incorporation into micelles or liposomes, transport across the brush border is extremely inefficient. Thus, both physiological and pharmaceutical solubilization methods for CoQ₁₀ (or CoQ₁₀H₂) are needed to enhance its absorption across the small intestine epithelium.

Small intestine epithelial cells express NPC1L1, a transport protein involved in cholesterol and vitamin E transport and metabolism (Figure 6a) [108,122,123]. Recent work has identified a role for NPC1L1 in the intestinal absorption of CoQ. NPC1L1 was overexpressed in Madin–Darby canine kidney (MDCK) epithelial cells to assess NPC1L1-mediated CoQ₉ and CoQ₁₀ transport [124]. MDCK cells are epithelial cells derived from kidney tissue of a cocker spaniel. NPC1L1 was expressed over 50-fold in human small intestine and duodenum compared to the kidney, making this mutant MDCK model reasonably analogous to that of intestinal CoQ uptake [125]. CoQ₁₀ and CoQ₉ uptake was found to increase significantly in NPC1L1-overexpressed MDCK cells, as compared to control MDCK cells. Furthermore, treatment with ezetimibe, a selective inhibitor of NPC1L1, decreased CoQ₁₀ and CoQ₉ uptake of NPC1L1-overexpressed cells back to control levels, implicating NPC1L1 as a CoQ-transport protein [124]. This study demonstrates the existence of intestine-specific proteins that function to increase CoQ₁₀ absorption, among other lipids. However, the mechanism of NPC1L1-CoQ₁₀ interaction remains unclear.

Figure 6.

CoQ₁₀ uptake and trafficking in mammalian cells. Dashed arrows indicate CoQ movement; solid arrows indicate chemical reactions. Proteins responsible for trafficking CoQ are identified where known; proteins mediating the transport shown by dashed arrows remain to be identified. (a) Exogenous CoQ uptake and intracellular CoQ distribution are mediated by several known mechanisms in mammalian cell culture. CD36 and NPC1L1 recognize and import exogenous CoQ within brown adipose tissue and small intestine epithelial cells, respectively. The endomembrane system facilitates intracellular trafficking of CoQ between membrane-bound vesicles. Dual-localized STARD7 in the mitochondria and cytosol support endogenous CoQ biosynthesis and distribution to the plasma membrane, respectively. (b) Model of lipoprotein-associated CoQ trafficking at the blood–brain barrier. SR-B1 and RAGE regulate influx of the HDL-associated CoQ across the blood–brain barrier, while LRP-1 regulates the efflux of LDL-associated CoQ. 1, receptor-mediated endocytosis; 2, recycling to apical side via the same receptor; 3, transcytosis to basolateral side; 4, transport to lysosome; 5, recycling or efflux of CoQ to apical membrane via LDL receptor family proteins; 6, paracellular transport detected due to defects in tight junctions. HDL, high-density lipoprotein; LDL, low-density lipoprotein. Created with BioRender.com (Accessed on 28 June 2023).

The HepG2 liver epithelial cell line has played an important role in the development of new strategies for exogenous CoQ₁₀ uptake. Moschetti et al. [126,127] engineered a nanodisk composed of a phosphatidylcholine bilayer circumscribed by apolipoprotein-A1, where CoQ₁₀ integrates readily within the midplane of the phosphatidylcholine bilayer. These nanodisk structures are defined as “reconstituted high-density lipoproteins” [128], and the CoQ₁₀ is integrated into a disk-shaped bilayer. The solubilization efficiency of the nanodisk-CoQ₁₀ was determined to be 97%, as compared to CoQ₁₀ which is essentially insoluble when dispersed in the presence of just apo-A1 or phosphatidylcholine [126]. The authors then assessed the disk’s ability to transport CoQ₁₀ into HepG2 cells. HepG2 cells exhibited a significant increase in mitochondrial CoQ₁₀ content upon incubation with nanodisk-associated CoQ₁₀ compared to empty nanodisks. The authors conclude this nanodisk to be a promising vehicle for exogenous CoQ₁₀ uptake and distribution in liver epithelium. However, the study lacked a control where CoQ₁₀ was supplemented in the absence of the nanodisk formation. This would have provided a useful point of comparison.

To further investigate the potential effectiveness of biomolecular nanodisks as a drug-delivery vehicle for CoQ₁₀, cultured C2C12 myotubes were treated with supra-pharmacological doses of simvastatin, an HMG-CoA reductase inhibitor that decreases the production of polyprenyl diphosphate precursors, and nanodisk-CoQ₁₀ [127]. Treatment with simvastatin caused a dose-dependent decrease in cell viability, reduced mitochondrial CoQ₁₀ content, and reduced the oxygen consumption rate. However, co-treatment with simvastatin and nanodisk-CoQ₁₀ increased mitochondrial CoQ₁₀ content compared to simvastatin treatment alone and ameliorated the reduction in oxygen consumption rate. These results highlight the potential of nanodisks to dramatically improve the bioavailability of CoQ₁₀. It would be interesting to assess the stability of CoQ₁₀H₂ in nanodisks and compare its uptake relative to that of the nanodisk-CoQ₁₀. Animal studies are needed to determine the efficacy of nanodisks as a drug-delivery vehicle in vivo.

3.1.2. Endothelial Cell Lines

Following absorption of CoQ₁₀ into the bloodstream, endothelial cells are the next point of uptake. Endothelial cells form the monolayer that line the luminal surface of blood vessels and capillaries, mediating macromolecule transport to and from neighboring tissues. The endothelium is studied extensively in the uptake and distribution of drugs due to this essential role in post-digestive tract drug transport. Human umbilical vein endothelial cells (HUVECs), human corneal endothelial cells, and bovine aortic endothelial cells (BAECs) have been utilized as models for CoQ uptake.

HUVECs have been used to study cell response to CoQ₁₀ as a protectant against oxidative stress. Treatment of HUVECs with Amyloid-β_1-42 or with a sub-fragment of Aβ_1–42 (Aβ_25–35) causes oxidative stress and elicits cell death via apoptosis and necrosis [129]. Pre-treatment with exogenous CoQ₁₀ (1–7 μM) was found to reduce reactive oxygen species (ROS) generation, inhibit the uptake of Aβ_25-35, and prevent cell death [130]. The results suggest that the uptake and trafficking of exogenous CoQ₁₀ is directed to sites essential in the quenching of free radicals within HUVECs [130].

Naguib et al. recently reported the uptake of CoQ₁₀H₂ in HCEC-B4G12 human corneal endothelial cells. HCEC-B4G12 cells exhibited a higher increase in total CoQ₁₀ content when supplemented with CoQ₁₀H₂/γ-cyclodextrin as compared to non-complexed CoQ₁₀ [120]. Furthermore, HCEC-B4G12 cells treated with ferroptosis-inducer erastin exhibited viability akin to that of untreated cells when incubated with as little as 1 µM CoQ₁₀H₂/γ-cyclodextrin, whereas HCEC-B4G12 cells treated with erastin and 100 µM non-complexed CoQ₁₀H₂ exhibited minimal rescue [120]. Uptake was also assessed by measuring ROS production in response to antimycin A, a mitochondrial electron transport chain Complex III inhibitor [120]. HCEC-B4G12 cells demonstrated a dose-dependent rescue of ROS generation in response to supplementation with CoQ₁₀H₂/γ-cyclodextrin, whereas non-complexed CoQ₁₀H₂ produced no significant change [120]. The authors report a promising level of CoQ₁₀H₂/γ-cyclodextrin uptake by this corneal endothelial cell line, pointing towards possible exogenous CoQ₁₀H₂ formulations involving γ-cyclodextrin to protect against oxidative stress-induced loss of function in human corneas.

BAECs incubated with non-complexed CoQ₁₀ do not exhibit significant levels of uptake. When BAECs were treated with glucose oxidase, an enzyme that oxidizes glucose to produce hydrogen peroxide, incubation with MitoQ reduced the activity of apoptotic enzymes Caspase-3 and Caspase-9, indicating rescue from oxidative stress [131]. Conversely, incubation with CoQ₁₀ did not reduce apoptotic activity, suggesting poor uptake of CoQ₁₀ and/or impaired trafficking to sites essential for protection against oxidative stress [131]. Furthermore, when BAECs pre-treated with 1 µM Mito-Q were incubated in either glucose oxidase or hydroperoxides, complex I activity (as measured via the rate of NADH oxidation) exhibited significant rescue compared to the untreated control; whereas pre-treatment with 1 µM CoQ₁₀ resulted in insignificant rescue in both glucose oxidase and hydroperoxide incubated BAECs [131]. Perhaps CoQ₁₀H₂/γ-cyclodextrin treatment could improve the uptake in BAECs, like that discussed for CaCo-2 and corneal epithelial cells.

3.1.3. Human Blood Cells

The HL-60 cell line is a human myeloid leukemia cell derived from a patient with promyelocytic leukemia. These cells may be cultured in suspension. The trafficking of endogenously synthesized ¹⁴C-labeled CoQ₁₀ was compared with the trafficking of exogenously added CoQ₉ [83]. HL-60 cells were incubated with ¹⁴C-labeled 4-HB, the ring precursor of CoQ biosynthesis (Figure 2), and the content of ¹⁴C-labeled CoQ₁₀ was determined in subcellular fractions over a period of 24 h. Fractions enriched in mitochondria contained more than 90% of the ¹⁴C-labeled CoQ₁₀, as ascertained at time points from two to 24 h. Fractions enriched in ER and mitochondria-associated membranes (MAM) contained small amounts of ¹⁴C-labeled CoQ₁₀ by two hours, and the plasma membrane fraction had detectable ¹⁴C-CoQ₁₀ only after six hours [83]. HL-60 cells treated with exogenous CoQ₉ showed an accumulation of CoQ₉ in mitochondrial and MAM fractions, followed by its subsequent accumulation in ER and plasma membrane fractions. After twelve hours of incubation with CoQ₉, the mitochondrial fraction contained the highest content of CoQ₉, followed by the ER and MAM fractions with similar CoQ₉ content, whereas the plasma membrane fraction exhibited significantly lower CoQ₉ content [83]. The endo-lysosomal membrane fraction was most enriched in CoQ₉, but this was not tracked during the time course. Treatment of HL-60 cells with brefeldin A, an inhibitor of vesicle formation and transport between the ER and Golgi, prior to the addition of ¹⁴C-4-HB resulted in a profound inhibition in the content of ¹⁴C-CoQ₁₀ in all fractions, with the plasma membrane fraction being most decreased. It is interesting that inhibition of trafficking between the ER and Golgi had such a dramatic effect on the endogenous synthesis of CoQ₁₀ within the mitochondria, and may indicate the importance of contact sites between the ER and mitochondria. Treatment of HL-60 cells with brefeldin A and CoQ₉ resulted in most of the CoQ₉ being sequestered in the endo-lysosomal fraction [83]. The high CoQ₉ content in the endo-lysosomal fraction suggests the importance of the endo-lysosomal trafficking pathway in the uptake and assimilation of exogenous CoQ₉. These findings indicate the importance of a functional endomembrane system in the synthesis of endogenous CoQ₁₀ and in the transport of both exogenous CoQ₉ and endogenous CoQ₁₀.

A recent paper by Wani et al. [132] found exogenous CoQ₁₀ (non-complexed) to ameliorate oxidative stress in erythrocytes and lymphocytes, suggesting its uptake. Erythrocytes and lymphocytes isolated from donor blood samples were exposed to TiO₂ nanoparticles to induce oxidative stress. Upon treatment with TiO₂, erythrocytes exhibited impaired ATPase activity and increased ROS production, and lymphocytes had severely depleted ATP levels and reduced mitochondrial membrane potential. Incubation with TiO₂ in the presence of CoQ₁₀ produced significant rescue of all markers of oxidative stress to wild-type or near-wild-type levels in both erythrocytes and lymphocytes [132]. These data indicate the ability of lymphocytes to take up and distribute exogenous CoQ₁₀ to the mitochondria, where it restored mitochondrial membrane potential, ATPase activity, and ATP levels. Lymphocytes are also able to traffic exogenous CoQ₁₀ to other cellular membranes, where it can slow lipid peroxidation induced by oxidative stress.

3.1.4. Human Dermal Cells

Human dermal fibroblasts (HDFs) are specialized cells found in the dermis that play a vital role in the maintenance and repair of the skin. The use of HDFs is especially relevant to investigate the efficacy of CoQ₁₀ as a commercially available skincare and anti-wrinkle product. HDFs exhibit significant levels of exogenous CoQ₁₀ uptake, as measured by a reduction in ROS generation [133]. The authors used pollutant particulate matter to activate KU812 cells, a basophilic leukocyte cell line, and used the resulting conditioned medium as a stress treatment of HDFs. Such treatment elicits a proinflammatory response and ROS generation in HDFs. Following eight hours of treatment, the HDFs were then treated with 100 nM CoQ₁₀. The exogenous CoQ₁₀ affected a significant reduction in mitochondrial ROS generation [133]. These findings suggest the ability of HDFs to take up and traffic CoQ₁₀ to the mitochondria.

HDFs have also been used to demonstrate an efficacious uptake of liposomal CoQ₁₀. CoQ₁₀ was suspended into liposomes via thin-film hydration, with a drug entrapment efficiency of 73.1% [134]. Administration of this liposomal-CoQ₁₀ suspension to HDFs that were oxidatively stressed from hydrogen peroxide treatment resulted in a complete rescue of cell viability to the point of cell proliferation, whereas non-complexed CoQ₁₀ administration of the same concentration caused no significant change in cell viability [134]. Similarly, quantification of ROS in these HDFs revealed that liposomal-CoQ₁₀ supplementation significantly reduced ROS production, while supplementation with non-complexed CoQ₁₀ produced no significant change [134]. The uptake of CoQ₁₀ by HDFs is thus augmented by liposomal suspension, where the CoQ₁₀ can restore antioxidant function.

HDFs exhibit improved uptake of exogenous CoQ₁₀H₂ versus CoQ₁₀. Incubation of HDFs with simvastatin decreased CoQ₁₀ levels by 29% [135]. A 24-h supplementation with 15 μg/mL CoQ₁₀H₂ caused a dose-dependent increase in total cellular CoQ₁₀ content, up to an 87-fold increase compared to the vehicle control. In contrast, supplementation with 15 μg/mL CoQ₁₀ resulted in just a 4.6-fold increase, a 95% reduction compared to CoQ₁₀H₂ [135]. Isolated mitochondria in HDFs treated with CoQ₁₀H₂ also exhibited a 160-fold increase in total CoQ₁₀ content, compared to a 2.5-fold increase with CoQ₁₀ treatment. Thus, treatment with CoQ₁₀H₂ resulted in enhanced mitochondrial CoQ₁₀ content relative to treatment with CoQ₁₀ [135]. The administration of CoQ₁₀H₂ to enhance uptake is promising, but the mechanisms that produce such high uptake relative to CoQ₁₀ remain uncharacterized.

Keratinocytes are epidermal skin cells that constitute the majority of the human epidermis, essential in forming the primary defense against environmental damage. CoQ₁₀ treatment has recently been shown to protect keratinocytes from oxidative stress-induced cell death, indicating its uptake [136]. HaCaT human keratinocytes derived from the foreskin were incubated with CoQ₁₀ directly added to Dulbecco’s modified Eagle’s Medium (DMEM) and fetal bovine serum (FBS), prior to treatment with the free-radical generating compound 2,2′-azobis-(2-amidinopropane) dihydrochloride (AAPH). Interestingly, no rescue of cell viability was observed until the DMEM-CoQ₁₀ medium was enhanced with CoQ₁₀ premixed in FBS, which resulted in a significant rescue of cell viability in the keratinocytes [136]. Similarly, intracellular CoQ₁₀H₂ content increased significantly after incubation with the premixed CoQ₁₀ medium, as compared to both the control and the original, insolubilized CoQ₁₀ medium [136]. The authors suggest that uptake of CoQ₁₀ by keratinocytes is augmented by the administration of CoQ₁₀ in a premixed medium, where otherwise an inefficient uptake is observed.

Keratinocytes have proved useful in the characterization of a new iodine-labeled CoQ derivative that may allow monitoring of the biodistribution of exogenous CoQ [137]. The novel iodine-labeled CoQ₁₀ (I₂-Q₁₀) consists of CoQ₁₀ with two iodine atoms replacing the terminal carbon of the polyisoprenyl tail. Iodine fluoresces upon excitation via X-rays, allowing for quantification and localization of I₂-Q₁₀ with minimal overlap between other molecules. Human keratinocytes incubated in 50 µM I₂-Q₁₀ showed no signs of toxicity or cell death, with levels of CoQ₁₀ uptake comparable to that of unlabeled CoQ₁₀. X-ray fluorescence imaging of pelleted keratinocytes incubated in I₂-Q₁₀ revealed a cell-specific localization of the iodine fluorescence with no extra-pellet iodine leakage. Furthermore, incubation with 50 µM I₂ resulted in a significant decrease in the pellet’s fluorescence, demonstrating poor uptake of elemental iodine and thus a mechanism for I₂-Q₁₀ uptake that is independent of iodine labeling uptake per se. The authors found exogenous I₂-Q₁₀ to distribute homogeneously over the keratinocyte cell, but co-localization with subcellular organelles, such as mitochondria, was not determined. Beyond the confirmation that CoQ₁₀ is taken up by human keratinocytes, these data suggest a potential way to assess CoQ₁₀ distribution using I₂-Q₁₀ in X-ray fluorescence imaging [137]. The technique was demonstrated to allow for quantitative analysis with no phenotypic changes due to the iodine labeling, and it shows promise in the investigation of mechanisms underlying exogenous CoQ uptake. It is also worth investigating whether the supplementation of with I₂-Q₁₀ restores respiratory function in a CoQ-deficient cell culture model, as some CoQ derivatives suffer from a loss of function upon modification of the tail (e.g., Mito-Q).

3.1.5. HeLa Cells

HeLa cells, derived from cervical cancer cells, are the oldest and most generally used human cell line in medicine to date. Recent work in HeLa cells has identified STARD7 as a phosphatidyl transfer protein involved in the export of endogenous CoQ₁₀ from the mitochondria to the plasma membrane, where it plays a role in ferroptosis suppression. STARD7 is localized to both the mitochondrial intermembrane space and the cytosol (Figure 6a). Deletion of STARD7 causes a nearly 5-fold increase in ferroptosis in response to treatment with PUFAs and erastin [63]. STARD7 is postulated to deliver mitochondrial CoQ₁₀ to the plasma membrane, where it is reduced by FSP1 to CoQ10₁₀H₂ and suppresses ferroptosis. Expression of a cytosol-localized form of STARD7 in HeLa cells significantly decreased cell death in response to treatment with erastin and PUFAs. Inhibition of either CoQ₁₀ biosynthesis or FSP1 function completely abolished resistance to ferroptosis, resulting in wildtype-levels of cell death in response to oxidative stress [63]. Expression of the mitochondria-localized form of STARD7 alone resulted in a significant decrease in plasma membrane-associated CoQ₁₀, while the mitochondrial CoQ₁₀ content was unaffected. The introduction of cytosol-localized STARD7 rescued plasma membrane-associated CoQ₁₀ content, suggesting that STARD7 functions as a CoQ₁₀ transporter in both the mitochondria and cytosol [63].

HeLa cells have also been used to study the effect of supplemented CoQ₁₀ on sulfide oxidation. SQR is the initial enzyme in the sulfide oxidation pathway, whose impairment can lead to increased ROS. CoQ₁₀ is an essential cofactor for SQR and protects the cell against oxidative stress. Silencing of COQ8A RNA in HeLa cells decreased the CoQ₁₀ content by approximately 50% and led to enhanced ROS production. Supplementation of CoQ₁₀ via Hydro-Q-Sorb decreased ROS production to near-control levels [138]. However, when SQR was knocked down in these CoQ₁₀-deficient cells, exogenous CoQ₁₀ no longer decreased ROS production [138]. This finding indicates that the supplementation of CoQ₁₀ rescued ROS levels due to CoQ₁₀-SQR activity, which requires CoQ₁₀ localization to the inner mitochondrial membrane. Thus, HeLa cells could take up the Hydro-Q-Sorb CoQ₁₀ and traffic it to the inner mitochondrial membrane, where it rescued sulfide oxidation.

HeLa cells have been utilized in the characterization of a lipopeptide-based surfactant, caspofungin (CF), that has been shown to solubilize CoQ. CF can form micellized nanoparticles with CoQ₁₀ up to 200 nm in diameter, with a critical micelle concentration of 50 µM [139]. Wang and Hekimi recently found that CF-solubilized CoQ₁₀ is a highly efficient formulation for exogenous supplementation in HeLa cells. Upon incubation for one hour, the CoQ₁₀ content in human HeLa cells increased significantly when treated with CoQ₁₀-CF as compared to CoQ₁₀, demonstrating the complex’s rapid uptake. Recent results further demonstrated that treatment with CoQ₁₀-CF restored maximal respiration in CoQ₁₀-deficient human fibroblasts derived from a patient harboring a homozygous mutation in the COQ7 gene [140]. The mechanism for CF micellization of CoQ₁₀ remains unknown, and the potential for CoQ₁₀-CF as a therapeutic to treat CoQ₁₀ deficiencies in patients warrants investigation.

3.1.6. Mouse Embryonic Fibroblasts (MEFs)

Wang et al. found that treatment with CoQ₁₀-CF can increase the uptake of exogenous CoQ₁₀ in MEFs in a dose-dependent manner. Incubation of MEFs for 72 h with 20 μM CF and 0.25 μM CoQ₁₀ produced a near-doubling in basal and maximal oxygen consumption rate as compared mouse embryonic fibroblasts incubated in just 10 μM CF without exogenous CoQ₁₀ [139]. After just a one-hour incubation with the CoQ₁₀-CF complex, MEF CoQ₁₀ content was approximately 3-fold higher when compared to MEFs incubated with free CoQ₁₀ for 48 h, indicating the rapid uptake of the complex.

3.2. Cell Receptors for Exogenous CoQ Uptake and Transport

3.2.1. Receptor-Mediated Uptake of CoQ

Receptor-mediated endocytosis plays a key role in the selective uptake of extracellular molecules, including CoQ₁₀. Wainwright et al. [141], used an in vitro blood–brain barrier (BBB) endothelial cell model of CoQ₁₀ deficiency, and identified lipoprotein-associated uptake and efflux mechanisms regulating CoQ₁₀ distribution across the BBB (Figure 6b). Porcine brain endothelial cells (PBECs) were grown in monolayer on a permeable membrane (see Figure 5), and transport of CoQ₁₀ was assessed from both the apical to basal (A to B, blood-to-brain side), and basal to apical (B to A, brain-to-blood side). The transport of non-complexed CoQ₁₀ across porcine brain endothelial cells (PBEC) was extremely low. Pre-treatment of CoQ₁₀ with bovine plasma-derived serum led to its association with the VLDL/LDL lipoprotein fractions and increased CoQ₁₀ transport across the PBEC monolayer by 4-fold, suggesting that CoQ₁₀ may be transported across the BBB as part of a lipoprotein complex. Transport of serum-treated CoQ₁₀ in the A to B direction matched transport in the B to A direction, indicating that opposing transport systems produced no net accumulation of CoQ₁₀ in the brain. This was observed for both the PBEC model and for a mouse BBB cell line, bEnd.3 [141]. Transporters at the BBB generally act together to limit systemic lipoprotein and cholesterol transfer to the brain.

Based on known lipoprotein transporters at the BBB, pharmacological inhibitors were screened for their effect on CoQ₁₀ transport across bEND.3 cells. BLT-1 is an inhibitor that targets the scavenger receptor (SR-B1), which interacts with HDL on the apical membrane to execute caveolae-mediated lipoprotein influx [142]. RAP was used to inhibit the low-density lipoprotein receptor (LDLR) family members, including the LDL receptor-related protein 1 (LRP-1). LRP-1 interacts with a variety of ligands to mediate BBB integrity [143]. FPS-ZM1 was used to inhibit the receptor for advanced glycation end products (RAGE). RAGE mediates the influx and opposes LRP-1-mediated efflux of ApoE and amyloid-beta across the BBB [144]. Upon inhibition of either SR-B1 or RAGE, apical-to-basal transport of CoQ₁₀ was decreased to 44% and 50% of the control, respectively, whereas basal-to-apical transport exhibited no change [141]. The authors suggested a direction-specific role of both receptors for CoQ₁₀ transcytosis into the brain. In contrast, inhibition of LRP-1 (and LDLR family members) increased apical-to-basal CoQ₁₀ transport by 168%, with no significant change in the reverse transport, suggesting that LRP-1 directs CoQ₁₀ transport out of the brain [141].

Wainwright et al. used pABA, a competitive inhibitor of polyprenyldiphosphate-4-hydroxybenzoate transferase (Coq2), to create a CoQ₁₀-deficient model of the BBB. Treatment with pABA significantly decreased cellular CoQ₁₀ content and mitochondrial respiratory chain complex activities in bEND.3 cells [141]. Cells treated with pABA also had reduced tight junction integrity and increased permeability (paracellular transport, Figure 6b), causing a shift in the net transport of CoQ₁₀ in the blood-to-brain direction. Under these pABA-inhibited conditions, the RAGE receptor was responsible for some of the net CoQ₁₀ transport, while the SR-B1 receptor-mediated transport was no longer active. However, the use of pABA to inhibit CoQ₁₀ biosynthesis has possible confounding effects. pABA is a substrate for Coq2, and the resulting 4-amino-3-polyprenylbenzoic acid may exert negative effects on cell endocytic trafficking (see Section 2.1). Studies with mammalian cultured cells show that once pABA is prenylated, it is converted to a redox-active “dead-end” amino-containing analog of DMQ, which may interfere with respiratory electron transport [47]. Instead of pABA, it has been suggested that 4-nitrobenzoic acid (4-NB) should be used as a competitive inhibitor of Coq2, since it is not prenylated [47].

Interestingly, co-supplementation of CoQ₁₀ with α-tocopherol or Trolox, a water-soluble analog of α-tocopherol that does not interact with lipoproteins, in pABA-treated and untreated bEND.3 cells caused an increase in CoQ₁₀ efflux in the brain-to-blood direction. This result suggests that the co-supplementation of CoQ₁₀ with α-tocopherol for individuals with CoQ₁₀ deficiency might decrease CoQ₁₀ delivery to the brain [141]. More work is needed to determine whether this effect is due to excessive antioxidants blocking uptake of lipoproteins containing CoQ₁₀ or if mechanisms of CoQ₁₀ uptake are mediated by changes in oxidative stress.

Finally, ABCB1 (P-glycoprotein) was studied since it is reported to decrease CoQ₁₀ transport across the CaCo-2 intestinal epithelial-barrier model [145]. Inhibition of P-glycoprotein with verapamil had no effect on CoQ₁₀ transport across the BBB, implying a mechanistic difference between exogenous CoQ₁₀ uptake in small intestinal epithelial cells versus BBB endothelial cells [141]. However, Wainwright et al. note that in the experiments by Itagaki et al. [145] with CaCo-2, exogenous CoQ₁₀ was supplemented in its free, un-micellized form, which is not favorable for small intestine uptake CoQ₁₀. The existence of brain endothelial membrane receptors that enhance CoQ₁₀ transcytosis in a selective direction across the BBB opens the possibility for antagonistic and/or agonistic supplements to increase exogenous CoQ₁₀ delivery to the brain.

CD36 is another receptor that has been identified in the uptake of exogenous CoQ. CD36 is a member of the SR family of transmembrane glycoproteins and a class-B scavenger receptor that exhibits high structural and functional homology to SR-B1 [146]. CD36 transports fatty acids and plays a regulatory role in lipid metabolism [147]. HEK293 human embryonic kidney cells overexpressing CD36 took up 57% more CoQ₉ than wild-type cells [148]. Importantly, receptor-mediated uptake of exogenous CoQ₉ and CoQ₁₀ was observed only when solubilized with Intralipid^®, an emulsion of soybean oil, glycerol, and egg yolk phospholipids. Receptor-mediated uptake was not reported for solvent- or detergent-solubilized CoQ [148]. Additional work by Anderson et al. [148] reports that CD36 drives CoQ uptake in brown adipose tissue (BAT) and is required for normal BAT function (see Section 4.1).

3.2.2. Saposin B Binds CoQ₁₀ and Regulates CoQ₁₀ Content

Prosaposin is a precursor protein that undergoes proteolytic cleavage to generate four small proteins known as saposins (saposins A, B, C, and D). Saposin B is a heat-stable glycoprotein that is detected in many tissues. It forms shell-like dimers to bind sphingolipids and glycerol-phospholipids, and primarily localizes to lysosomes [149,150,151]. CoQ₁₀ bound to saposin B was first identified in protein fractions collected from human urine samples and later detected in cell-lysates from HepG2 hepatocytes and human sperm [65]. CoQ₁₀ was shown to be transferred to erythrocyte ghost membranes when incubated with saposin B-CoQ₁₀ complexes at pH 4.5, suggesting that saposin B may play a role in delivery of CoQ₁₀ to acidic compartments, such as the lysosome [65].

Saposin B has high binding affinity for CoQ₁₀ as well as CoQ₉ and CoQ₇ isoforms, but a low affinity for α-tocopherol at pH 7.4 [65]. His-tagged saposin B expressed in E. coli was also shown to bind CoQ₁₀ in in vitro experiments in a pH-dependent manner, and showed that glycosylation of saposin B is not essential for CoQ₁₀ binding [152]. Saposin B has a comparable affinity for γ-tocopherol and CoQ₁₀ at neutral and basic conditions, but a higher affinity for γ-tocopherol at acidic conditions, suggesting that the binding affinity of saposin B for γ-tocopherol is greater than that of CoQ₁₀ [64]. In binding competition assays, two lipids in hexanes were mixed with aqueous saposin B at pH 7.4 and the resulting aqueous saposin B-bound lipids were measured. Interestingly, saposin B bound γ-tocopherol over CoQ₁₀ despite similar binding affinities and bound α-tocopherol over CoQ₁₀ despite lower binding affinity [64]. Based on known interactions of saposin B dimers with phospholipids, the authors postulated that saposin B dimers could bind two molecules of CoQ₁₀ or one molecule of CoQ₁₀ and two molecules of tocopherol. Their results suggest that the binding of one molecule of CoQ₁₀ may accelerate the binding of two molecules of α-tocopherol. The crystal structure of saposin B binding chloroquine was later resolved [153] and molecular docking studies identified two binding sites that are accessible to a variety of ligands [154].

Mutations in prosaposin and changes to its cellular content have been reported to affect CoQ₁₀ levels and mitochondrial function. Knockdown of prosaposin or overexpression of a prosaposin mutant that cannot form saposin B in HepG2 cells causes a significant reduction in both whole-cell and mitochondrial CoQ₁₀ content [155]. Likewise, overexpression of prosaposin leads to an increase in whole-cell and mitochondrial CoQ₁₀ content, suggesting that prosaposin and/or saposin B plays a role in regulating CoQ₁₀ levels [155]. Knockdown of prosaposin in CaCo-2 cells also decreases cellular CoQ₁₀ content, causing a decrease in ATP formation and glucose consumption [156]. As observed previously, a decrease in ATP formation is associated with disruption in tight junction formation and increased permeability.

More recent work has modeled long-term CoQ₁₀ deficiency in HepG2 cells and found that decreased CoQ₁₀ content was associated with decreased prosaposin levels [157]. Treatment with 4-NB decreased CoQ₁₀ cellular content after three days and decreased prosaposin protein levels after three months. Decreased CoQ₁₀ content and prosaposin protein levels were observed up to sixteen months with 4-NB treatment. The addition of the CoQ precursor 4-HB to cells treated with 4-NB significantly rescued both CoQ₁₀ content and prosaposin protein levels. Studies in prosaposin-deficient mice have not investigated CoQ content, though these mice often die one to two days after birth [158]. In summary, the disruption of prosaposin or saposin B decreases CoQ₁₀ content, indicating that saposin B is a CoQ₁₀ transporter that plays a role in regulating CoQ₁₀ levels.

4. Rodent Models

Mus musculus (mice), Rattus norvegicus (rats), and Cavia porcellus (guinea pigs) are common rodent models for the study of CoQ biosynthesis and uptake. Both mice and rats endogenously synthesize CoQ₉ as the predominant CoQ isoform, with CoQ₁₀ present as a minor isoform. This is in contrast to humans, guinea pigs, and other mammalian models, where CoQ₁₀ is the major isoform [8]. Rodent models are also used to study the effect of genetic mutations on endogenous CoQ biosynthesis [159]. Hence, rodent models have been used to assess how CoQ supplements affect the distribution and content of CoQ in tissues in both normal animals as well as those harboring gene mutations causing primary CoQ deficiency [6,160].

4.1. Organism-Level Uptake of CoQ

Due to their insolubility in water CoQ₉ and CoQ₁₀ are poorly bioavailable, necessitating unusually high doses [2,7]. Dosing in humans averages around 200 mg of CoQ₁₀ taken twice daily. However, when used to treat neurodegenerative diseases, doses can reach up to 3000 mg per day [161]. As a high-molecular-weight, lipophilic molecule, CoQ₁₀ is poorly absorbed by the gastrointestinal (GI) tract [162]. However, the consumption of CoQ₁₀ with other dietary lipids, as well as various supplement formulations, can enhance its uptake [7,162].

A majority of rodent studies investigate the effects of CoQ₁₀ supplements via oral treatment. The addition of CoQ₁₀ to rodent chow is a common feeding method. Similar to studies in human patients, high doses have been used to study CoQ₁₀ uptake in mice and rats [162,163]. As the CoQ₁₀-supplemented chow is digested, the hydrophobic CoQ₁₀ is taken up at the GI tract. In male Sprague Dawley rats, CoQ₁₀ permeability along the GI tract was investigated using a side-by-side diffusion apparatus, in which isolated GI tract tissues were placed in between a CoQ₁₀-containing donor medium and a receptor chamber [164]. These studies indicated that the greatest permeability is located at the duodenum, followed by the colon and the ileum.

Within the intestinal enterocytes, absorbed CoQ₁₀ is likely to be packaged along with other dietary lipids into chylomicrons, which are transported into the lymphatic system [165]. Chylomicrons in the lymph are then delivered to systemic circulation via the thoracic duct and subjected to processing by lipoprotein lipase in the capillaries. The resulting chylomicron remnants are taken up by the liver (Figure 7) [166]. This delivery may be responsible for CoQ₁₀ enrichment seen in the liver. We note here the assumption that orally supplemented CoQ₁₀ is trafficked in a manner similar to other dietary lipids such as vitamin E [167] and cholesterol [168]. However, there is a lack of direct experimental support for the presence of supplemented CoQ₁₀ in chylomicron remnants.

Figure 7.

Model for the absorption of dietary-supplemented CoQ in mice. Dashed arrows indicate CoQ distribution to tissues enriched only at high doses or IP injection. Stars in urine and feces indicate metabolites of CoQ. The ? indicates an unknown mechanism for the uptake of CoQ into the brain. VLDL—very low-density lipoprotein; IDL—intermediate-density lipoprotein; LDL—low-density lipoprotein; LPL—lipoprotein lipase. Created with BioRender.com (Accessed 31 January 2023).

In the liver, CoQ₁₀ derived from dietary sources is likely to be combined with de novo synthesized CoQ where it would be assembled into lipoprotein particles, with the majority in very low-density lipoprotein (VLDL) [164]. VLDL secreted by the liver is processed by lipoprotein lipase in the capillaries of various tissues to form intermediate-density lipoprotein and low-density lipoprotein (LDL) [168]. Lipoprotein particles extracted from human plasma find LDL to be most enriched in CoQ₁₀ content [169]. CoQ₁₀H₂ is an important antioxidant in circulating lipoproteins and enhances the resistance of LDL to lipid oxidation [170,171].

High-density lipoprotein particles (HDL) isolated from human plasma contain approximately 50% of the CoQ₁₀ content as compared to LDL [169]. This relationship between LDL and HDL CoQ content is also observed in mice [148]. In a mouse model, the absence of apolipoprotein ApoA1, a major component of HDL, decreased mitochondrial cardiac CoQ content by 75% [172]. Following ischemia/reperfusion of the left ascending coronary artery, the infarct size increased significantly in the homozygous ApoA1^−/− mice, as compared to the heterozygous ApoA1^−/+ mice, which in turn had larger infarcts than the parental control (WT) mice [172]. Although oral supplementation of CoQ₁₀ in ApoA1^−/− mice did not affect the cardiac CoQ₁₀ content, intraperitoneal (IP) injection of CoQ₁₀ restored the cardiac mitochondrial CoQ₁₀ content of the ApoA1^−/− mice to WT levels, and decreased their infarct size to that of WT mice following ischemia-reperfusion [172]. These results suggest a possible role for ApoA1 in the uptake and trafficking of CoQ₁₀ from the GI tract to the heart; however, further study is needed to confirm this relationship. It is unfortunate that in this study the authors did not quantify the endogenous CoQ₉ content as it may have further informed the outcome of supplementation with CoQ₁₀. However, another confounding factor is that the authors used CoQ₉ as an internal standard for the quantification of CoQ₁₀. This is problematic for two reasons: (1) mouse heart contains endogenous CoQ₉, and (2) commercially available standards of CoQ₉ can be contaminated with CoQ₁₀.

It is important to note that many studies in mice and rats utilize CoQ₁₀ as the supplemented isoform, although the predominant endogenous isoform is CoQ₉. This may generate conflicting outcomes when comparing rats or mice with guinea pigs, which solely produce CoQ₁₀ as the endogenous isoform. Several studies in both rats and mice have shown that supplementation of CoQ₁₀ increases endogenous CoQ₉ levels in various tissues, serum, and mitochondria [163,173,174]. This is likely due to a sparing effect of supplemented CoQ₁₀ on the endogenous CoQ₉ pool in which the addition of CoQ₁₀ to the cellular CoQ pool “spares” the turnover of CoQ₉; a similar phenomenon is observed when CoQ₁₀ supplements preserve levels of α-tocopherol or the tissue content of CoQ₉ [175,176]. The cardioprotective effects of CoQ₉ and CoQ₁₀ supplementation were investigated in guinea pigs [177]. CoQ₉ and CoQ₁₀ were found to have functionally identical effects, as both isoforms conferred comparable cardioprotection “as evidenced from the comparable degree of the postischemic ventricular recovery and reduction in myocardial infarct size and cardiomyocyte apoptosis”. Lekli et al. further claimed that CoQ₉ can be converted to CoQ₁₀ after supplementation, although no metabolic labeling of the fed CoQ₉ was performed to enable a more substantive conclusion. It is more likely the elevation in CoQ₁₀ content was due to the aforementioned sparing effect.

CoQ₁₀ present in lipoproteins may be delivered to cells via receptor-mediated uptake, to facilitate the entry of CoQ at the cellular level [2]. A specific example is the CD36 scavenger receptor-mediated uptake of exogenously supplied CoQ₁₀ into mouse brown adipose tissue (BAT) [148]. CD36 is present in many tissues, with the highest abundance in adipose tissue, heart, lung, and muscle, and less expression in spleen and liver [148]. BAT functions in non-shivering thermogenesis, a process generated by uncoupled respiration [178]. To perform this function, BAT depends on the uptake of CoQ rather than through its own de novo biosynthesis [148]. BAT isolated from mice lacking the CD36 receptor (CD36^−/−), have significantly reduced endogenous CoQ levels. Uptake of Intralipid^®-solubilized CoQ₁₀ in CD36^−/− mice via intraperitoneal injection is significantly reduced in BAT compared to wild-type mice, while no difference was observed in liver or muscle tissues. Moreover, CD36^−/− mice show a significant increase in CoQ₁₀ serum levels after intraperitoneal injection of exogenous CoQ₁₀ when compared to wild-type mice, suggesting that BAT is a major sink for exogenous CoQ. CD36^−/− BAT mitochondria showed impaired substrate utilization and decreased basal rates of respiration, which was rescued upon the addition of CoQ₂. CD36 is found to be exclusively on the plasma membrane in BAT, and BAT lacking CD36 lacks the characteristic browning of BAT following cold exposure. Likewise, CD36^−/− mice exhibit defective non-shivering thermogenesis. The mechanism by which CD36 mediates the uptake of CoQ₉ or CoQ₁₀ into BAT is not clear. The authors suggested that uptake could proceed either via CD36-mediated endocytosis of lipoprotein particles or that CoQ could be selectively taken up from a bound lipoprotein [148]. In either case, these findings provide evidence for receptor-mediated uptake of CoQ. (See Section 3.2.1 for further discussion of receptor-mediated uptake of CoQ₁₀ at the blood–brain barrier.)

Extensive reviews and literature compare the efficacy of supplementation with CoQ₁₀ and CoQ₁₀H₂ supplementation in humans [2,7,179]. However, few studies investigate the relative bioavailability of CoQ₁₀H₂ relative to CoQ₁₀ in a rodent model. In rats, CoQ₁₀ is converted into CoQ₁₀H₂ during or after absorption by the intestine [162], yet exogenous supplementation with CoQ₁₀H₂ has been posited to increase the molecule’s bioavailability. Mice lacking the ability to synthesize CoQ due to homozygous partial loss of function mutations in Coq9 (Coq9^−/−) were supplemented with water-soluble formulations of both CoQ₁₀ and CoQ₁₀H₂ [180]. While both formulations increased tissue levels relative to the Coq9^−/− control, the tissues of mice fed CoQ₁₀H₂ tended to have greater enrichment in CoQ₁₀ compared to mice fed CoQ₁₀ [180]. Measurement of cerebrum mitochondrial CoQ₁₀ content revealed the same trends. A recent study used positron emission tomography imaging to detect the uptake and distribution of ¹¹C-CoQ₁₀ and ¹¹C-CoQ₁₀H₂ that were administered to rats via tail vein injection [181]. The uptake of the labeled CoQ₁₀H₂ in the cerebrum, cerebellum, white adipose tissue, muscle, kidney, testis, and BAT was higher than that of the labeled CoQ₁₀. However, uptake of the ¹¹C-labeled CoQ₁₀H₂ was lower in the spleen as compared to labeled CoQ₁₀. Although the delivery of CoQ₁₀H₂ in this ¹¹C-labeling study was via intravenous injection [181], it is interesting that the enrichment achieved by CoQ₁₀H₂ in brain tissues was also observed for diet-delivered CoQ₁₀H₂ in the Coq9^−/− mouse study [180].

4.2. Tissue Localization of Supplemented CoQ

Understanding the localization of supplemented CoQ to specific tissues may be important to treating primary CoQ deficiencies that affect specific or multiple organs. However, as discussed above, the enrichment of CoQ in specific tissues of rats and mice can be affected by the redox state of the supplement. There is also variability observed in rats, mice, and guinea pigs following supplementation with different formulations of CoQ. This variability can be attributed to the conditions of supplementation including dose, mode of supplement administration, lipid extraction, and quantification methods used, as well as the genetic background or strain of the animal model.

4.2.1. Rat

IP injection into rats of ³H-CoQ₁₀, labeled at the isoprenyl methylene closest to the benzoquinone ring, permitted quick evaluation and extensive uptake into a number of tissues. Enriched tissues include the liver, spleen, adrenals, and ovaries, with limited uptake into the heart and thymus, as well as white blood cells [182]. This is in contrast to the poor dietary uptake of ³H-CoQ₁₀. Over a two-week period of oral supplementation, only 4–6% was taken up while the rest was excreted as radioactive metabolites in urine and feces. Analysis of the excreta revealed that half of the radioactive content excreted was present as non-metabolized CoQ₁₀ in feces, suggesting that excess CoQ₁₀ is removed through bile excretion (Figure 7) [182]. The results concur with prior studies that showed both oral and IP injections of CoQ₁₀ resulted in liver and spleen enrichment, while none was detected in the kidney, heart, or brain tissue [173,183]. However, Kwong et al. [174] observed modest CoQ₁₀ enrichment of the heart and kidney tissues. Differences in enrichment between studies may be attributed to dosage and method of feeding. Reahal and Wrigglesworth and Lonnrot et al. supplemented CoQ₁₀ in suspension [173,183], whereas Kwong et al. appear to supplement the chow with powdered CoQ₁₀ [174]. CoQ₁₀ is better taken up as an emulsion compared to a powder [2,7].

Intravenous (IV) administration in rats of 4 mg/kg CoQ₁₀ solubilized in lipid microspheres or chromophore EL has also been investigated [184]. Chromophore EL was described as a polyethoxylated castor oil that contained approximately 40 units of ethylene oxide to each triglyceride unit along with the addition of undefined antioxidants [185]. Results from Scalori et al. [184] indicate that CoQ₁₀, whether orally or IV administered, leaves the blood plasma quickly and is particularly enriched in the liver, although with varying kinetic profiles. The heart and kidneys of these rats were also enriched in CoQ₁₀. However, smaller doses of CoQ₁₀ were needed to achieve the same effect when supplemented by IV [184].

Topical application of CoQ₁₀, a lesser-studied method of administration, has also been found to increase skin concentrations in Sprague Dawley rats [184]. This is consistent with more recent data indicating that topical administration of CoQ₁₀ on the forearms of human females significantly increased levels at the skin surface and the deeper epidermis layer [186], and reduced UVB-induced wrinkle formation [187].

4.2.2. Mouse

Similar results for tissue enrichment of CoQ are seen in mouse models. Following a 13-week regimen of CoQ₁₀ supplementation in a soybean oil suspension, the serum and liver homogenate had increased levels of CoQ₁₀ [175]. No changes to CoQ₁₀ content were observed in the brain, heart, kidney, or skeletal muscle homogenates. However, a higher dose, 200 mg/kg/day, and a longer regimen of CoQ₁₀ were shown to increase cerebral cortex levels of CoQ₁₀, CoQ₁₀H₂, and CoQ₉H₂ [188]. These results are recapitulated in a study looking at the short-term administration of CoQ₁₀H₂ [189]. Imaging mass spectrometry also shows increased CoQ₁₀H₂ in several regions of the brain, including the hippocampus, cerebellum, and brain stem [189]. Matthews et al. showed oral supplementation of CoQ₁₀ exerted neuroprotective effects by reducing striated lesion volumes produced from systematic administration of 3-nitropropionic acid, a complex II inhibitor [188].

4.2.3. Guinea Pig

In the guinea pig model, Passi et al. [190] investigated the effect of CoQ₁₀ and ubiquinol₁₀-diacetate (CoQ₁₀-DIA) provided in the diet, on liver subcellular fraction content. CoQ₁₀H₂ modified by acetate at the ring 1,4 positions, (CoQ₁₀-DIA) is readily hydrolyzed and taken up as CoQ₁₀H₂. The study found that both CoQ₁₀ and CoQ₁₀-DIA taken up were mainly enriched in the heavy mitochondria and light mitochondria/lysosome fractions. Modest increases in CoQ₁₀ content were also seen in the nucleus/debris and microsome fractions. There were no significant differences in subcellular enrichment between CoQ₁₀ and CoQ₁₀-DIA [190].

Across all three rodent models, exogenously supplemented CoQ₁₀ is enriched mainly in the liver, plasma, and serum. While there remains conflicting data on CoQ enrichment in other tissues, it appears that high dosage, IV, or IP administration may be able to overcome certain barriers to uptake into tissues such as the brain, heart, and kidney.

4.3. Subcellular Localization of Supplemented CoQ

When the CoQ sequestered in lipoproteins is taken up by a cell, it must be trafficked to intracellular membranes. Currently, little is known about the mechanisms of intracellular trafficking used to shuttle CoQ. However, the subcellular enrichment of CoQ following supplementation has been studied in mice, rats, and guinea pigs.

4.3.1. Rat

Under normal physiological conditions, endogenous CoQ₉ in rat liver tissue is predominantly located in the mitochondrial subcellular fraction [183]. After CoQ₁₀ supplementation, subcellular fractionation of hepatic tissue cells was investigated [191]. The pellet containing lysosomes and light mitochondria was especially enriched in CoQ₁₀, with no change to the endogenous levels of CoQ₉. Bentinger et al. [182] conducted an investigation into the subcellular localization of ³H-CoQ₁₀ in the liver, finding that nearly 60% of radiolabeled CoQ₁₀ localized to the lysosome after IP supplementation, whereas the mitochondrial fractions contained only approximately 11% of the total ³H-CoQ₁₀ recovered [182]. This is consistent with a prior study, where IP administered CoQ₁₀ was primarily localized to the soluble cytosolic fraction of liver cells and was recovered as the oxidized form, not the reduced form [183]. As CoQ₁₀ is highly hydrophobic, it is unclear why the concentration was found to be greatest in the soluble cytosolic fraction. It is also worth mentioning that this study does not indicate precautions taken to mitigate autoxidation of reduced CoQ₁₀H₂ in liver extracts during tissue homogenization and lipid extraction. Such precautions may include quick procedural timing, the use of acidified solvents during extraction, and inert gases to prevent autoxidation. An LC–MS/MS quantification method sensitive to the CoQ redox state was recently developed to minimize measurement error due to autoxidation [119].

In a separate rat study, the brain and brain mitochondria were shown to be enriched in CoQ₁₀ following a two-month supplementation regimen at a dose of 200 mg CoQ₁₀/kg per day [188]. However, the CoQ₉ content was not significantly increased in this study. Another study administering 150 mg/kg CoQ₁₀ per day for 13 weeks demonstrated an increase in heart, skeletal muscle, liver, and kidney mitochondrial CoQ₁₀ content, as well as an additional increase in CoQ₉ content in mitochondria of heart, skeletal muscle, and kidney tissue homogenate [174]. Only the mitochondria-enriched fractions were examined for CoQ₉ and CoQ₁₀ contents in these two studies.

4.3.2. Mouse

Mice supplemented with CoQ₁₀ generally presented with increases in both CoQ₉ and CoQ₁₀ content in the liver, heart, kidney, and skeletal muscle tissue mitochondria, but not in the brain [163,176]. The data obtained by Sohal et al. were subjected to further analyses to consider the length of treatment and dosage factors. General trends indicated that mice treated for 17.5 months with CoQ₁₀ had larger magnitude increases in CoQ content as compared to mice treated for 3.5 months [163]. No consistent dose-dependent trend was reported. However, these conclusions are solely based on the examination of mitochondria-enriched fractions

4.3.3. Guinea Pig

In the guinea pig model, Passi et al. [190] investigated the effect of dietary CoQ₁₀ and CoQ₁₀-DIA on liver subcellular fraction content. The authors found that both CoQ₁₀ and CoQ₁₀-DIA taken up were mainly enriched in the heavy mitochondria and light mitochondria/lysosome fractions. Modest increases in CoQ₁₀ content were also seen in the nucleus/debris and microsome fractions. There were no significant differences in subcellular enrichment between CoQ₁₀ and CoQ₁₀-DIA.

4.4. Intracellular Trafficking of CoQ

There is little information on the intracellular trafficking of exogenously supplemented or endogenously synthesized CoQ in rodent models. Data gleaned from other cellular models, including Saccharomyces cerevisiae and human cell lines, can inform such studies. Intriguingly, ADCK2, the mouse homolog of yeast Cqd1 (discussed in Section 2.3), has been shown to play a role in modulating the transport of lipids into mitochondria for fatty acid β-oxidation and CoQ biosynthesis [95]. ADCK2 and Cqd1 are members of the UbiB/Coq8 family of atypical kinases. Adck2^+/− mice exhibit haploinsufficiency, with CoQ deficiency specifically in skeletal muscle and defects in mitochondrial fatty acid oxidation [95]. CoQ₁₀ supplementation increased the content of CoQ₁₀ in the skeletal muscle of the Adck2^+/− mice, enhanced physical performance, and decreased plasma lactate levels [95]. It seems likely that ADCK2 acts similarly to Cqd1, the yeast ortholog, and is necessary to retain CoQ within mitochondria [93].

4.5. Mouse Models of CoQ Deficiency and the Effects of CoQ Supplementation

Investigation of human mutations in COQ genes has been studied in transgenic mouse models to assess the organism-level effect of canonical single nucleotide variants. Transgenic mice with homologous gene KOs have also been used to assess and investigate phenotypic consequences, and to evaluate the efficacy of treatment with CoQ₁₀ supplements.

Saiki et al. [192] investigated the efficacy of CoQ₁₀ supplementation in the Pdss2^kd/kd (kidney disease) mouse. The Pdss2^kd/kd mouse harbors a spontaneous homozygous V117M mutation resulting in a partial loss of function in PDSS2, one of the two subunits of polyprenyldiphosphate synthase essential for CoQ biosynthesis, and a homolog of yeast Coq1 [192,193]. Pdss2^kd/kd mice present with CoQ deficiency in early adulthood, resulting in a kidney disease similar to the collapsing glomerulopathy variant of focal segmental glomerulosclerosis [192]. Mice harboring a podocyte-specific loss of function knockout of Pdss2 reproduced a more severe phenotype of the Pdss2^kd/kd mice, indicating that the kidney disease in the Pdss2^kd/kd mouse results from the CoQ deficiency in the podocytes [193]. Intriguingly, supplementation with CoQ₁₀ afforded a dramatic rescue of the kidney disease as ascertained by a 5-fold reduction in urine albumin content. However, CoQ₁₀ supplementation, even at doses of 200 mg/kg per day did not affect the kidney CoQ₁₀ content as compared to the untreated Pdss2^kd/kd mouse. It was speculated that the CoQ₁₀ supplementation may exert its effects via the filtration of blood plasma components [194]. The podocyte foot processes act as a filtration barrier, and the CoQ₁₀ may function primarily as a membrane antioxidant [192]. A more recent study using the same Pdss2^kd/kd mouse model supports CoQ₁₀ supplementation (0.5% CoQ₁₀ supplemented food), as a preventative treatment for the progression of nephrotic syndrome induced by deficiency [138]. Kidney CoQ₉ and CoQ₁₀ levels were increased, and damage indicated by renal histology is significantly dampened after six months of supplementation with CoQ₁₀. CoQ₁₀ supplementation rescued defects in the oxidation of sulfide, and decreased oxidative stress in the kidneys of Pdss2^kd/kd mice [138].

The kidneys of Pdss2^kd/kd mice are shown to have a defect in the sulfide oxidation pathway resulting in increased hydrogen sulfide and depleted glutathione [195]. Deficiencies in CoQ impede the sulfide oxidation pathway by decreasing SQR activity [195,196]. Further investigation of mouse Coq9 mutants revealed similar results. Coq9^R239X mice, with 10-15% residual CoQ₉ content, had significantly less SQR levels in the kidneys and cerebrum whereas Coq9^Q95X, with 40–50% residual CoQ₉ content in the same tissues, had no decrease in these tissues [196]. Both models exhibit 10–20% residual CoQ₉ content in muscle and have decreased SQR in this tissue. While low CoQ content disrupts sulfide oxidation, supplementation with 100 μM CoQ₁₀ upregulates the sulfide oxidation pathway, restoring its function as well as other downstream and tangential effects of impaired sulfur metabolism [197].

Investigation of the mouse Coq7 homolog, Mclk1, showed that homozygous knockout resulted in embryonic lethality [198]. Heterozygous Mclk1^+/− mice, however, live a normal lifespan with no significant defects in subcellular CoQ₉ levels or mitochondrial morphology. Subcellular distribution of CoQ₉ is altered, which may explain the lowered oxygen consumption and ATP production phenotype of heterozygous Mclk1^+/− mitochondria [199,200]. Supplementation with CoQ₁₀ significantly increases mitochondrial CoQ₁₀ content and restores the electron transport chain defect [198].

Rare deleterious mutations in the human COQ9 gene result in severe multisystem disorders, affecting the central nervous system, cardiac, and renal systems [201]. Homozygous Coq9 mice (Coq9^R239X/R239X) were prepared to model the human COQ9^R244X deficiency [180]. These mice exhibit profound brain damage due to oxidative damage, neuronal apoptosis, and demyelination in the pons and medulla oblongata [180]. After two months of 240 mg/kg/day CoQ₁₀ or CoQ₁₀H₂ oral supplementation, significant increases were detected in plasma, liver, and muscle tissue, followed by smaller increases in the heart, kidney, cerebrum, and cerebellum [180]. Supplementation with CoQ₁₀H₂ appeared to alleviate histopathological symptoms of CoQ deficiency in the brain, whereas CoQ₁₀ did so to a lesser extent.

Such dramatic effects of CoQ₁₀ supplementation have also been seen in some human patients harboring primary CoQ deficiency [202,203,204,205]. The parallels between the rodent disease models and the human clinical outcomes emphasize the benefit of CoQ-deficient mouse research models.

5. Conclusions

CoQ is an essential and ubiquitous lipid with many complex roles. However, there remains much to be learned about the mechanisms responsible for its uptake, intracellular trafficking, and distribution. Here, we have summarized the work in several model organisms used to investigate how CoQ is trafficked.

Yeast is an extraordinarily powerful model organism that has been used to identify proteins involved in the uptake and trafficking of exogenous CoQ₆ to mitochondrial respiratory complexes. The yeast coq mutants completely lack CoQ₆ biosynthesis and are thus respiratory deficient and fail to grow in a medium containing a non-fermentable carbon source. Supplementation with exogenous CoQ restores respiration and growth of coq mutants, and supplementation with the soluble CoQ₂ analog can bypass defects in endocytosis and membrane trafficking [18,19,58]. Fernandez-del-Rio et al. [18] identified six genes that are essential for the uptake and trafficking of exogenous CoQ₆ to mitochondrial respiratory complexes—RTS1, CDC10, RVS161, RVS167, VPS1, and NAT3. More work is needed to determine the role that each of these genes plays in the uptake and trafficking of exogenous CoQ₆. While supplementation of ORFΔcoq2Δ double mutants with exogenous CoQ₆ is a powerful approach for identifying genes required for CoQ₆ uptake and trafficking, genes that are redundant or essential for respiration cannot be identified with exogenous CoQ₆ rescue screens.

Yeast is also a powerful model for determining pathways and proteins responsible for the trafficking of endogenously synthesized CoQ₆. Biosynthesis and transport of endogenously synthesized CoQ₆ from the mitochondria in yeast is possibly mediated by ERMES [88,89]. However, direct transport of CoQ₆ through ERMES has not yet been observed. Cqd1 and Cqd2 are mitochondrial inner membrane proteins that face the intermembrane space and influence the cellular distribution of endogenous CoQ₆, and likely play a role in transporting CoQ₆ from the mitochondrial inner membrane to the outer membrane and ER [93]. It is likely that other still unidentified CoQ₆ transporters may contribute to trafficking.

Mammalian cell lines are an ideal model for investigating the cellular uptake and intracellular distribution of CoQ. Synthetic vehicles, such as cyclodextrin, caspofungin, micelles, liposomes, and nanodisks have been shown to dramatically improve the bioavailability of exogenous CoQ [117]. Moreover, supplementation with CoQ₁₀H₂ displays improved bioavailability in human dermal fibroblasts compared to CoQ₁₀ [135], a trend that is also observed in rodent models [180,181]. Treatment of HL-60 cells with brefeldin-A shows that intracellular distribution of both endogenous and exogenous CoQ relies at least in part on the endomembrane system [83], though more work is needed to identify essential players of this pathway for CoQ transport. Recent work in HeLa cells has identified STARD7 as a phosphatidyl transfer protein involved in the export of endogenous CoQ from the mitochondria to the plasma membrane, where it plays a role in ferroptosis suppression [63]. Work in mouse endothelial brain cells has identified scavenger receptor SR-B1, receptor for advanced glycation endproducts (RAGE), and LRP-1 to be involved in transport of lipoprotein-associated CoQ₁₀ across the blood–brain barrier [141]. CoQ₁₀ may also be delivered to BAT cells by receptor-mediated uptake, via-the CD36 scavenger receptor [2,148].

Rodent models are useful for investigating the bioavailability of CoQ supplements and CoQ distribution in tissues of both wild-type animals as well as those harboring gene mutations that cause primary CoQ₁₀ deficiency [6,160]. Dietary CoQ in rodents is taken up at the GI tract [164], and is then likely packaged into chylomicrons by intestinal enterocytes and transported into the lymphatic system, though more work needs to be performed to identify the pathways of endothelial uptake and to detect the supplemented CoQ in chylomicrons. CoQ from chylomicron remnants is likely repackaged with de novo CoQ at the liver and secreted in lipoprotein particles, with the majority in VLDL. Tissue localization of supplemented CoQ in rodents varies across studies and CoQ formulations, but enrichment is usually observed in the liver, plasma, and serum, with varying results in other tissues such as the heart, brain, and kidney. Supplemented CoQ is enriched in subcellular mitochondrial and light mitochondrial/lysosomal fractions. Moreover, transgenic mouse models have been used to investigate human mutations and canonical single nucleotide variants in Coq genes in order to evaluate the efficacy of treatment of CoQ-deficient mouse models with CoQ₁₀ supplements.

It is important to note that comparisons across studies are challenging due to variability in CoQ formulations and dosage. Comparative studies of CoQ₁₀ and CoQ₁₀H₂ must use proper precautions and controls to prevent rapid oxidation of CoQ₁₀H₂. Unfortunately, rodent studies that investigate CoQ₁₀ uptake sometimes overlook endogenous CoQ₉ content, despite CoQ₉ being the predominant isoform in mice and rats. Many aspects of CoQ uptake and transport remain unknown or poorly understood. What is the relative importance of the endomembrane system and membrane contact sites in the transport of CoQ? How are CoQ transport and intracellular distribution of CoQ regulated? Elucidating the pathways of how organisms assimilate CoQ supplements, its cellular uptake, and pathways of intracellular distribution is important to understanding the functional roles of CoQ and may lead to improved therapeutics for exogenous supplementation.

Abbreviations

BAECs, bovine aortic endothelial cells; BAT, brown adipose tissue; BBB, blood–brain barrier; CoQ, coenzyme Q or ubiquinone; CoQH, ubisemiquinone; CoQH₂, ubiquinol; CoQn, coenzyme Qn where n describes the number of isoprene units in the polyisoprenyl tail; COQ, denotes yeast or human gene required for biosynthesis of CoQ; Coq, denotes rodent genes; coq, denotes yeast gene harboring a mutation; COQ, denotes human or rodent polypeptide; Coq, denotes yeast polypeptide; CHDH, choline dehydrogenase; cyb5Red, NADH-cytochrome b₅ reductase; DDMQH₂, 2-methoxy-6-polyprenyl-1,4-benzohydroquinone; DHOD, dihydroorate dehydrogenase; DHP, 4,5-dihydroxy-3-polyprenylphenol; DHPB, 4,5-dihydroxy-3-polyprenylbenzoic acid; DMAPP, dimethylallyl pyrophosphate; DMQH₂, 2-methoxy-5-methyl-6-polyprenyl-1,4-benzohydroquinone; DMeQH₂, 3-methyl-6-methoxy-2-polyprenyl-1,4,5-benzenetriol; DMGDH, dimethylglycine dehydrogenase; ERMES, endoplasmic reticulum encounter structure; ETF, electron transfer flavoprotein; ETFDH, electron transfer flavoprotein dehydrogenase; FSP1, ferroptosis suppressor protein 1; GPDH, glycerol-3-phosphate dehydrogenase; GPX4, glutathione peroxidase 4; HAB, 3-hexaprenyl-4-aminobenzoic acid; 4-HB, 4-hydroxybenzoic acid; 4-Hbz, 4-hydroxybenzaldehyde; HDL, high-density lipoprotein; 4-HMA, 4-hydroxymandelate; HMPB, 4-hydroxy-5-methoxy-3-polyprenylbenzoic acid; HPB, 4-hydroxy-3-polyprenyl-benzoic acid; 4-HPP, 4-hydroxyphenylpyruvate; HUVECs, human umbilical vein endothelial cells; IPP, isopentenyl pyrophosphate; LDL, low-density lipoprotein; LRP-1, LDL receptor related protein 1; MEFs, mouse embryonic fibroblasts; MICOS, mitochondrial contact site and cristae organization system; NPC1L1, Niemann–Pick C1-like 1; NQO1, NAD(P)H:quinone oxidoreductase 1; pABA, para-aminobenzoic acid; PBEC, porcine brain endothelial cells; PMRS, plasma membrane redox system; PRDH, proline dehydrogenase; PUFAs, polyunsaturated fatty acids; RAGE, receptor for advanced glycation end products; SARDH, sarcosine dehydrogenase; SQR, sulfide:quinone oxidoreductase; SR-B1, scavenger receptor; VLDL, very low-density lipoprotein.

Author Contributions

Writing—original draft preparation, A.J., K.A.A. and M.D.G.; writing—review and editing, A.J., C.F.C., K.A.A. and M.D.G.; visualization, A.J., C.F.C., K.A.A. and M.D.G.; supervision, C.F.C. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Funding Statement

This research received no external funding.